![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F2926708%2F3c9164a0d2caa56d9e7cb4c3b7fd3185%2FData%20competition%20social%20card%202.png?generation=1608128905290766&alt=media\")
Hi Kagglers!
\nThe CDP: Unlocking Climate Solutions Kaggle Competition has come to its conclusion and it is time to announce our prize winners!
\nThank you again to all those who submitted. It was incredibly tough to pick out a set of winners from all the Notebooks and our judges would like to congratulate all those who entered, it's great to see so many innovative solutions as we all work together to build towards a truly sustainable economy.
\nWe would like to thank all of you who have entered prize submissions. The range of data sets were fantastic and we can’t wait to see how you utilise the data to build further metrics as the competition continues.
\n1st place ($50,000)
\nWinning Submission: CDP Challenge Climate Adaptation Index
\nWinning Users: @shabou
In this R notebook, @shabou uses a thorough literature review and deep dive into domain knowledge, leading to adoption of an ND-GAIN inspired methodology to generate vulnerability and readiness KPIs for every city in the CDP dataset. Interactive maps and engaging visualisations then classify cities within a vulnerability/readiness matrix to target City level engagement and policy decisions, while accounting for key social indicators such as health system and potable water supply sensitivity to climate change risks.
\n2nd place ($25,000)
\nSubmission: A Path to Efficient and Sustainable Growth
\nUsers: @katemelianova ; @artvolgin
In this R notebook, @katemelianova and @artvolgin employed a sound mathematical technique of Data Envelopment Analysis (DEA) to analyze possible city efficiency for maximising KPI performance. This framework was then combined with elements of graph theory to identify where ‘Lagger’ Cities can bridge the gap towards similar Cities that are leading the way on environmental and social metrics. This approach aligned perfectly to prominent features of the CDP policy framework and suggested real world solutions, pinpointing specific policy areas where Cities can improve on both environmental and social equity metrics. Ensuring Corporations were not left behind, the analysis also identified where Cities and Corporations overlap in their identification of climate-related opportunities, applying Structural Topic Modelling (STM) of free text responses to assess topic similarity.
\n3rd place ($10,000)
\nSubmission: Impact Potential Analysis of Water Use Efficiency
\nUsers: @iamleonie
In this Water-focussed Python notebook, @iamleonie utilises distance metrics between two dimensions to create a complex impact potential KPI, weighting water use efficiency by the geospatial context a company operates within. This analysis elevates the role of water stressed geographies to a company’s KPI performance and leverages external datasets from WRIs Aquaduct tool .The notebook then uses these metrics brilliantly to rank countries, cities, corporations, and facilities according to both their performance and their potential for improvement.
\nSubmission: KPIs for Measuring Climate Action and Inequality
\nUsers: @mannmann2 ; @rebeccaverghese
In this comprehensive Python notebook, @mannmann2 and @rebeccaverghese create an extensive set of KPIs using the CDP dataset to explore a range of different environmental and economic issues.
\nSubmission:Tree Canopy Equality
\nUsers: @adriansarstedt ; @hamishgunasekara
This Python notebook takes an innovative approach to the challenge by looking at tree canopy equality within cities and how that perpetuates existing imbalances between the rich and the poor.
\nSubmission: Do You Have a Climate Plan?
\nUsers: @jbrans
In this Python notebook, @jbrans develops a number of different KPIs and visualises combined KPI scores for every city in the CDP dataset in an interactive map. KPI scores are plotted against “percent of population living in poverty” to intelligently explore the intersection between environmental issues and social/economic issues.
\nSubmission: CDP Microlevel Urban Sustainability Approach
\nUsers: @reubencpereira
In this Python notebook @reubencpereira utilized radar plots brilliantly to visualise their calculated CDP sustainability index, presenting a well-automated solution that could be scaled to many Cities across the CDP sample.
\nSubmission: You Can’t Manage What You Don’t Measure
\nUsers: @vlarmet
In this Python notebook, @vlarmet develops a sustainability ambition KPI and then maps the scores for every city in the CDP dataset. There is also an extensive narrative that discusses the intersection between environmental issues and social issues.
\nWell done to all our winners!
\nWe hope you enjoyed the competition and look forward to implementing some of your brilliant ideas within CDP.
\nHave a great holiday period and stay safe.
\nRegards,
\nThe CDP Team
", "messages": [ { "id": 1115758, "postDate": "2020-12-16T14:56:58.307Z", "content": "
Hi Kagglers!
\nThe CDP: Unlocking Climate Solutions Kaggle Competition has come to its conclusion and it is time to announce our prize winners!
\nThank you again to all those who submitted. It was incredibly tough to pick out a set of winners from all the Notebooks and our judges would like to congratulate all those who entered, it's great to see so many innovative solutions as we all work together to build towards a truly sustainable economy.
\nWe would like to thank all of you who have entered prize submissions. The range of data sets were fantastic and we can’t wait to see how you utilise the data to build further metrics as the competition continues.
\n1st place ($50,000)
\nWinning Submission: CDP Challenge Climate Adaptation Index
\nWinning Users: @shabou
In this R notebook, @shabou uses a thorough literature review and deep dive into domain knowledge, leading to adoption of an ND-GAIN inspired methodology to generate vulnerability and readiness KPIs for every city in the CDP dataset. Interactive maps and engaging visualisations then classify cities within a vulnerability/readiness matrix to target City level engagement and policy decisions, while accounting for key social indicators such as health system and potable water supply sensitivity to climate change risks.
\n2nd place ($25,000)
\nSubmission: A Path to Efficient and Sustainable Growth
\nUsers: @katemelianova ; @artvolgin
In this R notebook, @katemelianova and @artvolgin employed a sound mathematical technique of Data Envelopment Analysis (DEA) to analyze possible city efficiency for maximising KPI performance. This framework was then combined with elements of graph theory to identify where ‘Lagger’ Cities can bridge the gap towards similar Cities that are leading the way on environmental and social metrics. This approach aligned perfectly to prominent features of the CDP policy framework and suggested real world solutions, pinpointing specific policy areas where Cities can improve on both environmental and social equity metrics. Ensuring Corporations were not left behind, the analysis also identified where Cities and Corporations overlap in their identification of climate-related opportunities, applying Structural Topic Modelling (STM) of free text responses to assess topic similarity.
\n3rd place ($10,000)
\nSubmission: Impact Potential Analysis of Water Use Efficiency
\nUsers: @iamleonie
In this Water-focussed Python notebook, @iamleonie utilises distance metrics between two dimensions to create a complex impact potential KPI, weighting water use efficiency by the geospatial context a company operates within. This analysis elevates the role of water stressed geographies to a company’s KPI performance and leverages external datasets from WRIs Aquaduct tool .The notebook then uses these metrics brilliantly to rank countries, cities, corporations, and facilities according to both their performance and their potential for improvement.
\nSubmission: KPIs for Measuring Climate Action and Inequality
\nUsers: @mannmann2 ; @rebeccaverghese
In this comprehensive Python notebook, @mannmann2 and @rebeccaverghese create an extensive set of KPIs using the CDP dataset to explore a range of different environmental and economic issues.
\nSubmission:Tree Canopy Equality
\nUsers: @adriansarstedt ; @hamishgunasekara
This Python notebook takes an innovative approach to the challenge by looking at tree canopy equality within cities and how that perpetuates existing imbalances between the rich and the poor.
\nSubmission: Do You Have a Climate Plan?
\nUsers: @jbrans
In this Python notebook, @jbrans develops a number of different KPIs and visualises combined KPI scores for every city in the CDP dataset in an interactive map. KPI scores are plotted against “percent of population living in poverty” to intelligently explore the intersection between environmental issues and social/economic issues.
\nSubmission: CDP Microlevel Urban Sustainability Approach
\nUsers: @reubencpereira
In this Python notebook @reubencpereira utilized radar plots brilliantly to visualise their calculated CDP sustainability index, presenting a well-automated solution that could be scaled to many Cities across the CDP sample.
\nSubmission: You Can’t Manage What You Don’t Measure
\nUsers: @vlarmet
In this Python notebook, @vlarmet develops a sustainability ambition KPI and then maps the scores for every city in the CDP dataset. There is also an extensive narrative that discusses the intersection between environmental issues and social issues.
\nWell done to all our winners!
\nWe hope you enjoyed the competition and look forward to implementing some of your brilliant ideas within CDP.
\nHave a great holiday period and stay safe.
\nRegards,
\nThe CDP Team
", "rawMarkdown": "\n\n**Hi Kagglers!**\n\nThe [CDP: Unlocking Climate Solutions Kaggle Competition](https://www.kaggle.com/c/cdp-unlocking-climate-solutions/overview) has come to its conclusion and it is time to announce our prize winners!\n\nThank you again to all those who submitted. It was incredibly tough to pick out a set of winners from all the Notebooks and our judges would like to congratulate all those who entered, it's great to see so many innovative solutions as we all work together to build towards a truly sustainable economy.\n\nWe would like to thank all of you who have entered prize submissions. The range of data sets were fantastic and we can’t wait to see how you utilise the data to build further metrics as the competition continues.\n\n## Prize Winners\n\n**1st place ($50,000)**\n**Winning Submission:** [CDP Challenge Climate Adaptation Index ](https://www.kaggle.com/shabou/cdp-challenge-climate-adaptation-index)\n**Winning Users:** @shabou \n\nIn this R notebook, @shabou uses a thorough literature review and deep dive into domain knowledge, leading to adoption of an ND-GAIN inspired methodology to generate vulnerability and readiness KPIs for every city in the CDP dataset. Interactive maps and engaging visualisations then classify cities within a vulnerability/readiness matrix to target City level engagement and policy decisions, while accounting for key social indicators such as health system and potable water supply sensitivity to climate change risks.\n\n**2nd place ($25,000)**\n**Submission:** [A Path to Efficient and Sustainable Growth](https://www.kaggle.com/katemelianova/cdp-a-path-to-efficient-and-sustainable-growth) \n**Users:** @katemelianova ; @artvolgin \n\nIn this R notebook, @katemelianova and @artvolgin employed a sound mathematical technique of Data Envelopment Analysis (DEA) to analyze possible city efficiency for maximising KPI performance. This framework was then combined with elements of graph theory to identify where ‘Lagger’ Cities can bridge the gap towards similar Cities that are leading the way on environmental and social metrics. This approach aligned perfectly to prominent features of the CDP policy framework and suggested real world solutions, pinpointing specific policy areas where Cities can improve on both environmental and social equity metrics. Ensuring Corporations were not left behind, the analysis also identified where Cities and Corporations overlap in their identification of climate-related opportunities, applying Structural Topic Modelling (STM) of free text responses to assess topic similarity.\n\n**3rd place ($10,000)**\n**Submission:** [Impact Potential Analysis of Water Use Efficiency ](https://www.kaggle.com/iamleonie/impact-potential-analysis-of-water-use-efficiency)\n**Users:** @iamleonie \n\nIn this Water-focussed Python notebook, @iamleonie utilises distance metrics between two dimensions to create a complex impact potential KPI, weighting water use efficiency by the geospatial context a company operates within. This analysis elevates the role of water stressed geographies to a company’s KPI performance and leverages external datasets from [WRIs Aquaduct tool](https://www.wri.org/aqueduct) .The notebook then uses these metrics brilliantly to rank countries, cities, corporations, and facilities according to both their performance and their potential for improvement. \n\n## Honourable Mentions\n\n**Submission:** [KPIs for Measuring Climate Action and Inequality](https://www.kaggle.com/mannmann2/kpis-for-measuring-climate-action-and-inequality)\n**Users:** @mannmann2 ; @rebeccaverghese \n\nIn this comprehensive Python notebook, @mannmann2 and @rebeccaverghese create an extensive set of KPIs using the CDP dataset to explore a range of different environmental and economic issues.\n\n\n**Submission:**[Tree Canopy Equality](https://www.kaggle.com/adriansarstedt/cdp-tree-canopy-equality-kpi-final-submission)\n**Users:** @adriansarstedt ; @hamishgunasekara \n\nThis Python notebook takes an innovative approach to the challenge by looking at tree canopy equality within cities and how that perpetuates existing imbalances between the rich and the poor.\n\n**Submission:** [Do You Have a Climate Plan?](https://www.kaggle.com/jbrans/do-you-have-a-climate-plan)\n**Users:** @jbrans \n\nIn this Python notebook, @jbrans develops a number of different KPIs and visualises combined KPI scores for every city in the CDP dataset in an interactive map. KPI scores are plotted against “percent of population living in poverty” to intelligently explore the intersection between environmental issues and social/economic issues.\n\n**Submission:** [CDP Microlevel Urban Sustainability Approach](https://www.kaggle.com/reubencpereira/cdp-microlevel-urban-sustainability-approach)\n**Users:** @reubencpereira \n\nIn this Python notebook @reubencpereira utilized radar plots brilliantly to visualise their calculated CDP sustainability index, presenting a well-automated solution that could be scaled to many Cities across the CDP sample.\n\n**Submission:** [You Can’t Manage What You Don’t Measure](https://www.kaggle.com/vlarmet/cdp-you-can-t-manage-what-you-don-t-measure/)\n**Users:** @vlarmet \n\nIn this Python notebook, @vlarmet develops a sustainability ambition KPI and then maps the scores for every city in the CDP dataset. There is also an extensive narrative that discusses the intersection between environmental issues and social issues.\n\n**Well done to all our winners!**\n\nWe hope you enjoyed the competition and look forward to implementing some of your brilliant ideas within CDP. \n\nHave a great holiday period and stay safe.\n\nRegards,\n\n*The CDP Team*\n\n", "votes": 15 }, { "id": 1116006, "postDate": "2020-12-16T18:25:10.500Z", "content": "Congrats! Will we have feedbacks? I would like to have some please to improve my works.
", "rawMarkdown": "Congrats! Will we have feedbacks? I would like to have some please to improve my works.", "votes": 5, "replies": [ { "id": 1117012, "postDate": "2020-12-17T16:28:39.347Z", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant. Submissions were evaluated on accuracy/completeness, communication, and documentation.
\nFrom my perspective your submission looks great, although I might note that the narrative text sections appear to be less detailed than the winning submissions, and these narrative text sections are critically important for communicating to the reader exactly what you are doing, why you are doing it, and what was the significance of your result, for example.
", "rawMarkdown": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant. Submissions were [evaluated](https://www.kaggle.com/c/cdp-unlocking-climate-solutions/overview/evaluation) on accuracy/completeness, communication, and documentation. \n\nFrom my perspective [your submission](https://www.kaggle.com/boltmaud/started-with-dataviz-next-step-is-data-mining) looks great, although I might note that the narrative text sections appear to be less detailed than the winning submissions, and these narrative text sections are critically important for communicating to the reader exactly what you are doing, why you are doing it, and what was the significance of your result, for example. " } ] }, { "id": 1115914, "postDate": "2020-12-16T17:10:20.637Z", "content": "Congratulations to all winners!
\nTanks a lot to the competition organizers for offering the opportunity to work on such an interesting and impactful subject. It was a great experience for me working with CDP datasets!
\nThanks also to the participants who shared their notebooks, insights and datasets!
", "rawMarkdown": "Congratulations to all winners!\n\nTanks a lot to the competition organizers for offering the opportunity to work on such an interesting and impactful subject. It was a great experience for me working with CDP datasets!\n\nThanks also to the participants who shared their notebooks, insights and datasets!", "votes": 4 }, { "id": 1115768, "postDate": "2020-12-16T15:07:04.123Z", "content": "Congrats to the winners!
\n@shabou, @katemelianova, @artvolgin, and @iamleonie the Kaggle Team will reach out to you shortly with instructions on how to claim your prizes.
\nThere were a lot of great submissions, thank you everybody for making submissions and for marking your work as public!
", "rawMarkdown": "**Congrats to the winners!**\n\n@shabou, @katemelianova, @artvolgin, and @iamleonie the Kaggle Team will reach out to you shortly with instructions on how to claim your prizes.\n\nThere were a lot of great submissions, thank you everybody for making submissions and for marking your work as public! \n\n", "votes": 3, "replies": [ { "id": 1116013, "postDate": "2020-12-16T18:32:06.153Z", "content": "As an additional token of our appreciation we will also be sending @shabou, @katemelianova, @artvolgin, @iamleonie, @mannmann2, @rebeccaverghese, @adriansarstedt, @hamishgunasekara, @jbrans, @reubencpereira, and @vlarmet a Kaggle-branded t-shirt or hat.
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1314380%2Fee9fbc15ea8288e63057b4205ed4b6ca%2FScreen%20Shot%202020-12-16%20at%2011.18.11%20AM.png?generation=1608142819323879&alt=media\")
Great work everyone, and thank you for making your notebooks public!
", "rawMarkdown": "As an additional token of our appreciation we will also be sending @shabou, @katemelianova, @artvolgin, @iamleonie, @mannmann2, @rebeccaverghese, @adriansarstedt, @hamishgunasekara, @jbrans, @reubencpereira, and @vlarmet a Kaggle-branded t-shirt or hat. \n\n\n\nGreat work everyone, and thank you for making your notebooks public!", "votes": 3 } ] }, { "id": 1116637, "postDate": "2020-12-17T10:58:13.683Z", "content": "Congratulations to all winners!
\nWe are thankful to the organisers for holding this competition and offering such rich data for the analysis. Great job to the prize winners and all the participants!
", "rawMarkdown": "Congratulations to all winners!\n\nWe are thankful to the organisers for holding this competition and offering such rich data for the analysis. Great job to the prize winners and all the participants!\n", "votes": 2 }, { "id": 1116040, "postDate": "2020-12-16T19:24:08.120Z", "content": "@callumr22 @paultimothymooney Thank you very much for the recognition. I feel very honored and had a lot of fun with this challenge. Looking forward to the next one!
\nCongratulations @shabou @katemelianova @artvolgin to your outstanding work! Great job, guys!
\nAlso, great job to everyone who has participated. This was an interesting challenge given the amount of data and its structure. There were a lot of great notebooks!
", "rawMarkdown": "@callumr22 @paultimothymooney Thank you very much for the recognition. I feel very honored and had a lot of fun with this challenge. Looking forward to the next one! \n\nCongratulations @shabou @katemelianova @artvolgin to your outstanding work! Great job, guys!\n\nAlso, great job to everyone who has participated. This was an interesting challenge given the amount of data and its structure. There were a lot of great notebooks!", "votes": 2 }, { "id": 1115920, "postDate": "2020-12-16T17:13:28.687Z", "rawMarkdown": "", "votes": 4, "isDeleted": true, "replies": [ { "id": 1116062, "postDate": "2020-12-16T19:45:12.743Z", "content": "We would really appreciate a response. Thank you.
", "rawMarkdown": "We would really appreciate a response. Thank you.", "votes": 1 }, { "id": 1117006, "postDate": "2020-12-17T16:22:59.620Z", "content": "Decisions were made by a judging committee and all decisions are final. Your notebook was recognized with an honorable mention and you were awarded with a secondary prize but unfortunately only one submission can win first prize.
\nThanks again for participating and contributing a public notebook!
", "rawMarkdown": "Decisions were made by a judging committee and all decisions are final. Your notebook was recognized with an honorable mention and you were awarded with a secondary prize but unfortunately only one submission can win first prize. \n\nThanks again for participating and contributing a public notebook!" }, { "id": 1117128, "postDate": "2020-12-17T18:37:53.227Z", "content": "Thank you, Paul, for connecting back.
\nWhile I understand that only some of the participants can win prizes, many of us joined this competition cause of our passion for the Climate cause, and since we have given our ideas, I feel there should be a fair attempt at providing open and sound feedback to the participants. This is especially necessary in the case of an analytics competition such as this, where evaluation criterias are subjective in nature. In addition, we are also looking to understand what made the winning notebook stand out since there were multiple submissions (including ours) that were similar to or inspired by the ND-GAIN approach to vulnerability KPIs.
\nWe believe that your criteria on \"accuracy/completeness, communication, documentation\" was satisfied by our notebook. We would therefore be extremely grateful to you if the judging panel can be asked to make an exception and provide more insight into the way they evaluated these notebooks. In our case, if not for the rest of the 9 KPIs, we ask that we at least get feedback on the KPIs regarding Vulnerability, Adaptability/Readiness, and Collaboration, since these are the ones that are also present in the winning notebook.
\nThank you for your time.
", "rawMarkdown": "Thank you, Paul, for connecting back.\n\nWhile I understand that only some of the participants can win prizes, many of us joined this competition cause of our passion for the Climate cause, and since we have given our ideas, I feel there should be a fair attempt at providing open and sound feedback to the participants. This is especially necessary in the case of an analytics competition such as this, where evaluation criterias are subjective in nature. In addition, we are also looking to understand what made the winning notebook stand out since there were multiple submissions (including ours) that were similar to or inspired by the ND-GAIN approach to vulnerability KPIs.\n\nWe believe that your criteria on *\"accuracy/completeness, communication, documentation\"* was satisfied by our notebook. We would therefore be extremely grateful to you if the judging panel can be asked to make an exception and provide more insight into the way they evaluated these notebooks. In our case, if not for the rest of the 9 KPIs, we ask that we at least get feedback on the KPIs regarding Vulnerability, Adaptability/Readiness, and Collaboration, since these are the ones that are also present in the winning notebook.\n\nThank you for your time.", "votes": 2 }, { "id": 1117640, "postDate": "2020-12-18T09:32:16.130Z", "content": "Hi,
\nI acknowledge that judging notebooks is extremely subjective and I am not here to say that my notebook should have made it in the list above. I can see everyone has put loads of work an effort into those submissions as much if not more than I did, providing original ideas and analysis, so well done to you all who made it onto this discussion.
Having said this, I think the point of transparency plagues these analytics competitions, as no real feedback is provided and it I think it wouldn't be hard to return a score for each notebook, this has been raised many times before by others so maybe not worth discussing again and when I signed up I knew this was the process anyway.
\nLooking back, what really have left me a bit upset this time I must say, is the fact that I am the one who brought forward the ND-GAIN dataset for the Mid-Prize submission.
\nI am not from this field and it really took me ages to find it. That submission did not even make it in the honourable mentions at the Mid-Prize announcement.
\nI didn't really say anything at the time as I am no expert and, even if I thought it was super relevant, I trusted the judgment that was provided.
\nNow I see that ND-GAIN is not only the centrepiece of the winning notebook, but has been use in many if not most of the submissions above.
\nAt this point I think it can be a bit less subjective that this was indeed the most useful dataset as opposed to the ones which got recognised which were not as nearly used in the submissions who made it to the finalist shortlist.
\nAll this to say that this gives even less confidence in this quite opaque process and I at this stage I am really not sure if these assessments are produced in a consistent manner.
\nI enjoy Kaggle for the learning experience and its wonderful community and contributors, but form now on I plan to continue to use the website for modelling competitions only.
\nThis is the last time I will participate to an analytics competition, at least until the process becomes more transparent.
\nIs a bit of a shame because I was already getting started on another which is running right now on a similar topic, but I honestly concluded is really not worth the effort.
\nThe evaluation of the most useful dataset submission, which in my opinion becomes a bit less subjective given the importance in the winning submission and others above, really doesn't give me confidence in the way judgments are expressed.
\nRegards
", "rawMarkdown": "Hi,\nI acknowledge that judging notebooks is extremely subjective and I am not here to say that my notebook should have made it in the list above. I can see everyone has put loads of work an effort into those submissions as much if not more than I did, providing original ideas and analysis, so well done to you all who made it onto this discussion. \n\nHaving said this, I think the point of transparency plagues these analytics competitions, as no real feedback is provided and it I think it wouldn't be hard to return a score for each notebook, this has been raised many times before by others so maybe not worth discussing again and when I signed up I knew this was the process anyway.\n\nLooking back, what really have left me a bit upset this time I must say, is the fact that I am the one who brought forward the ND-GAIN dataset for the Mid-Prize submission. \n\nI am not from this field and it really took me ages to find it. That submission did not even make it in the honourable mentions at the Mid-Prize announcement.\n\nI didn't really say anything at the time as I am no expert and, even if I thought it was super relevant, I trusted the judgment that was provided.\n \nNow I see that ND-GAIN is not only the centrepiece of the winning notebook, but has been use in many if not most of the submissions above.\n\nAt this point I think it can be a bit less subjective that this was indeed the most useful dataset as opposed to the ones which got recognised which were not as nearly used in the submissions who made it to the finalist shortlist.\n\nAll this to say that this gives even less confidence in this quite opaque process and I at this stage I am really not sure if these assessments are produced in a consistent manner.\n\nI enjoy Kaggle for the learning experience and its wonderful community and contributors, but form now on I plan to continue to use the website for modelling competitions only. \n\nThis is the last time I will participate to an analytics competition, at least until the process becomes more transparent.\n\nIs a bit of a shame because I was already getting started on another which is running right now on a similar topic, but I honestly concluded is really not worth the effort.\n\nThe evaluation of the most useful dataset submission, which in my opinion becomes a bit less subjective given the importance in the winning submission and others above, really doesn't give me confidence in the way judgments are expressed.\n\nRegards\n\n ", "votes": 7 }, { "id": 1118129, "postDate": "2020-12-18T18:50:48.820Z", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant.
\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The USA-only and the Global ND-GAIN datasets are both great.
", "rawMarkdown": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant. \n\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The [USA-only](https://www.kaggle.com/david289/ndgain-uaa-dataset) and the [Global](https://www.kaggle.com/shabou/ndgain-country-index) ND-GAIN datasets are both great. \n" }, { "id": 1118130, "postDate": "2020-12-18T18:50:48.820Z", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant.
\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The USA-only and the Global ND-GAIN datasets are both great.
", "rawMarkdown": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant. \n\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The [USA-only](https://www.kaggle.com/david289/ndgain-uaa-dataset) and the [Global](https://www.kaggle.com/shabou/ndgain-country-index) ND-GAIN datasets are both great. \n" }, { "id": 1118198, "postDate": "2020-12-18T20:30:17.913Z", "content": "Paul, I do appreciate your prompt reply. I can see you got the point I was trying to make.
\nI just would like to challenge you further on this \"the judging committee is not able to provide detailed feedback\".. I disagree with this. Here in Kaggle you have such a big community of bright minds, why not running a thread or even a challenge in finding ways to improve the user experience in analytics competition?
\nI am sure once again machine learning could come in our aid and help you in selecting the shortlist of candidates, while highlighting what are the obvious faults of notebooks with \"no chance\" that a user might not be able to see.
\nAnd for the rest say top 20-50 for example a more detailed scorecard with feedback that could help people to improve for the next competition could be achievable.
I can see in every single analytics competition you always need to deal negative feedback and complaints, there is definitely a pattern here.
\nThis is just a thought, I think is in everyone's interest (including especially people like you who need to answer to each and single one of us) to improve user experience for this type of competitions.
", "rawMarkdown": "Paul, I do appreciate your prompt reply. I can see you got the point I was trying to make.\n\nI just would like to challenge you further on this \"the judging committee is not able to provide detailed feedback\".. I disagree with this. Here in Kaggle you have such a big community of bright minds, why not running a thread or even a challenge in finding ways to improve the user experience in analytics competition? \nI am sure once again machine learning could come in our aid and help you in selecting the shortlist of candidates, while highlighting what are the obvious faults of notebooks with \"no chance\" that a user might not be able to see. \nAnd for the rest say top 20-50 for example a more detailed scorecard with feedback that could help people to improve for the next competition could be achievable.\n\nI can see in every single analytics competition you always need to deal negative feedback and complaints, there is definitely a pattern here.\n \nThis is just a thought, I think is in everyone's interest (including especially people like you who need to answer to each and single one of us) to improve user experience for this type of competitions.", "votes": 4 } ] }, { "id": 1115821, "postDate": "2020-12-16T15:42:19.730Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 1116006, "author_name": "BoltMaud", "author_url": "", "post_date": "2020-12-16T18:25:10.500000", "content": "Congrats! Will we have feedbacks? I would like to have some please to improve my works.
", "votes": 5, "replies": [ { "id": 1117012, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-17T16:28:39.347000", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant. Submissions were evaluated on accuracy/completeness, communication, and documentation.
\nFrom my perspective your submission looks great, although I might note that the narrative text sections appear to be less detailed than the winning submissions, and these narrative text sections are critically important for communicating to the reader exactly what you are doing, why you are doing it, and what was the significance of your result, for example.
", "votes": 0, "replies": [] } ] }, { "id": 1115914, "author_name": "shabou", "author_url": "", "post_date": "2020-12-16T17:10:20.637000", "content": "Congratulations to all winners!
\nTanks a lot to the competition organizers for offering the opportunity to work on such an interesting and impactful subject. It was a great experience for me working with CDP datasets!
\nThanks also to the participants who shared their notebooks, insights and datasets!
", "votes": 4, "replies": [] }, { "id": 1115768, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-16T15:07:04.123000", "content": "Congrats to the winners!
\n@shabou, @katemelianova, @artvolgin, and @iamleonie the Kaggle Team will reach out to you shortly with instructions on how to claim your prizes.
\nThere were a lot of great submissions, thank you everybody for making submissions and for marking your work as public!
", "votes": 3, "replies": [ { "id": 1116013, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-16T18:32:06.153000", "content": "As an additional token of our appreciation we will also be sending @shabou, @katemelianova, @artvolgin, @iamleonie, @mannmann2, @rebeccaverghese, @adriansarstedt, @hamishgunasekara, @jbrans, @reubencpereira, and @vlarmet a Kaggle-branded t-shirt or hat.
\n
Great work everyone, and thank you for making your notebooks public!
", "votes": 3, "replies": [] } ] }, { "id": 1116637, "author_name": "Ekaterina Melianova", "author_url": "", "post_date": "2020-12-17T10:58:13.683000", "content": "Congratulations to all winners!
\nWe are thankful to the organisers for holding this competition and offering such rich data for the analysis. Great job to the prize winners and all the participants!
", "votes": 2, "replies": [] }, { "id": 1116040, "author_name": "Leonie", "author_url": "", "post_date": "2020-12-16T19:24:08.120000", "content": "@callumr22 @paultimothymooney Thank you very much for the recognition. I feel very honored and had a lot of fun with this challenge. Looking forward to the next one!
\nCongratulations @shabou @katemelianova @artvolgin to your outstanding work! Great job, guys!
\nAlso, great job to everyone who has participated. This was an interesting challenge given the amount of data and its structure. There were a lot of great notebooks!
", "votes": 2, "replies": [] }, { "id": 1115920, "author_name": "", "author_url": "", "post_date": "2020-12-16T17:13:28.687000", "content": "", "votes": 4, "replies": [ { "id": 1116062, "author_name": "Arman", "author_url": "", "post_date": "2020-12-16T19:45:12.743000", "content": "We would really appreciate a response. Thank you.
", "votes": 1, "replies": [] }, { "id": 1117006, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-17T16:22:59.620000", "content": "Decisions were made by a judging committee and all decisions are final. Your notebook was recognized with an honorable mention and you were awarded with a secondary prize but unfortunately only one submission can win first prize.
\nThanks again for participating and contributing a public notebook!
", "votes": 0, "replies": [] }, { "id": 1117128, "author_name": "Rebecca Verghese", "author_url": "", "post_date": "2020-12-17T18:37:53.227000", "content": "Thank you, Paul, for connecting back.
\nWhile I understand that only some of the participants can win prizes, many of us joined this competition cause of our passion for the Climate cause, and since we have given our ideas, I feel there should be a fair attempt at providing open and sound feedback to the participants. This is especially necessary in the case of an analytics competition such as this, where evaluation criterias are subjective in nature. In addition, we are also looking to understand what made the winning notebook stand out since there were multiple submissions (including ours) that were similar to or inspired by the ND-GAIN approach to vulnerability KPIs.
\nWe believe that your criteria on \"accuracy/completeness, communication, documentation\" was satisfied by our notebook. We would therefore be extremely grateful to you if the judging panel can be asked to make an exception and provide more insight into the way they evaluated these notebooks. In our case, if not for the rest of the 9 KPIs, we ask that we at least get feedback on the KPIs regarding Vulnerability, Adaptability/Readiness, and Collaboration, since these are the ones that are also present in the winning notebook.
\nThank you for your time.
", "votes": 2, "replies": [] }, { "id": 1117640, "author_name": "David 289", "author_url": "", "post_date": "2020-12-18T09:32:16.130000", "content": "Hi,
\nI acknowledge that judging notebooks is extremely subjective and I am not here to say that my notebook should have made it in the list above. I can see everyone has put loads of work an effort into those submissions as much if not more than I did, providing original ideas and analysis, so well done to you all who made it onto this discussion.
Having said this, I think the point of transparency plagues these analytics competitions, as no real feedback is provided and it I think it wouldn't be hard to return a score for each notebook, this has been raised many times before by others so maybe not worth discussing again and when I signed up I knew this was the process anyway.
\nLooking back, what really have left me a bit upset this time I must say, is the fact that I am the one who brought forward the ND-GAIN dataset for the Mid-Prize submission.
\nI am not from this field and it really took me ages to find it. That submission did not even make it in the honourable mentions at the Mid-Prize announcement.
\nI didn't really say anything at the time as I am no expert and, even if I thought it was super relevant, I trusted the judgment that was provided.
\nNow I see that ND-GAIN is not only the centrepiece of the winning notebook, but has been use in many if not most of the submissions above.
\nAt this point I think it can be a bit less subjective that this was indeed the most useful dataset as opposed to the ones which got recognised which were not as nearly used in the submissions who made it to the finalist shortlist.
\nAll this to say that this gives even less confidence in this quite opaque process and I at this stage I am really not sure if these assessments are produced in a consistent manner.
\nI enjoy Kaggle for the learning experience and its wonderful community and contributors, but form now on I plan to continue to use the website for modelling competitions only.
\nThis is the last time I will participate to an analytics competition, at least until the process becomes more transparent.
\nIs a bit of a shame because I was already getting started on another which is running right now on a similar topic, but I honestly concluded is really not worth the effort.
\nThe evaluation of the most useful dataset submission, which in my opinion becomes a bit less subjective given the importance in the winning submission and others above, really doesn't give me confidence in the way judgments are expressed.
\nRegards
", "votes": 7, "replies": [] }, { "id": 1118129, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-18T18:50:48.820000", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant.
\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The USA-only and the Global ND-GAIN datasets are both great.
", "votes": 0, "replies": [] }, { "id": 1118130, "author_name": "Paul Mooney", "author_url": "", "post_date": "2020-12-18T18:50:48.820000", "content": "There were a lot of submissions and unfortunately the judging committee is not able to provide detailed feedback to every participant.
\nAs per the midpoint prize, it does not appear as if any of the winning submissions used your dataset that looked at ND-GAIN values for US cities -- but we agree it is a great asset to have on the Kaggle platform and we thank you for sharing it! The USA-only and the Global ND-GAIN datasets are both great.
", "votes": 0, "replies": [] }, { "id": 1118198, "author_name": "David 289", "author_url": "", "post_date": "2020-12-18T20:30:17.913000", "content": "Paul, I do appreciate your prompt reply. I can see you got the point I was trying to make.
\nI just would like to challenge you further on this \"the judging committee is not able to provide detailed feedback\".. I disagree with this. Here in Kaggle you have such a big community of bright minds, why not running a thread or even a challenge in finding ways to improve the user experience in analytics competition?
\nI am sure once again machine learning could come in our aid and help you in selecting the shortlist of candidates, while highlighting what are the obvious faults of notebooks with \"no chance\" that a user might not be able to see.
\nAnd for the rest say top 20-50 for example a more detailed scorecard with feedback that could help people to improve for the next competition could be achievable.
I can see in every single analytics competition you always need to deal negative feedback and complaints, there is definitely a pattern here.
\nThis is just a thought, I think is in everyone's interest (including especially people like you who need to answer to each and single one of us) to improve user experience for this type of competitions.
", "votes": 4, "replies": [] } ] }, { "id": 1115821, "author_name": "", "author_url": "", "post_date": "2020-12-16T15:42:19.730000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1115758": "\n\n**Hi Kagglers!**\n\nThe [CDP: Unlocking Climate Solutions Kaggle Competition](https://www.kaggle.com/c/cdp-unlocking-climate-solutions/overview) has come to its conclusion and it is time to announce our prize winners!\n\nThank you again to all those who submitted. It was incredibly tough to pick out a set of winners from all the Notebooks and our judges would like to congratulate all those who entered, it's great to see so many innovative solutions as we all work together to build towards a truly sustainable economy.\n\nWe would like to thank all of you who have entered prize submissions. The range of data sets were fantastic and we can’t wait to see how you utilise the data to build further metrics as the competition continues.\n\n## Prize Winners\n\n**1st place ($50,000)**\n**Winning Submission:** [CDP Challenge Climate Adaptation Index ](https://www.kaggle.com/shabou/cdp-challenge-climate-adaptation-index)\n**Winning Users:** @shabou \n\nIn this R notebook, @shabou uses a thorough literature review and deep dive into domain knowledge, leading to adoption of an ND-GAIN inspired methodology to generate vulnerability and readiness KPIs for every city in the CDP dataset. Interactive maps and engaging visualisations then classify cities within a vulnerability/readiness matrix to target City level engagement and policy decisions, while accounting for key social indicators such as health system and potable water supply sensitivity to climate change risks.\n\n**2nd place ($25,000)**\n**Submission:** [A Path to Efficient and Sustainable Growth](https://www.kaggle.com/katemelianova/cdp-a-path-to-efficient-and-sustainable-growth) \n**Users:** @katemelianova ; @artvolgin \n\nIn this R notebook, @katemelianova and @artvolgin employed a sound mathematical technique of Data Envelopment Analysis (DEA) to analyze possible city efficiency for maximising KPI performance. This framework was then combined with elements of graph theory to identify where ‘Lagger’ Cities can bridge the gap towards similar Cities that are leading the way on environmental and social metrics. This approach aligned perfectly to prominent features of the CDP policy framework and suggested real world solutions, pinpointing specific policy areas where Cities can improve on both environmental and social equity metrics. Ensuring Corporations were not left behind, the analysis also identified where Cities and Corporations overlap in their identification of climate-related opportunities, applying Structural Topic Modelling (STM) of free text responses to assess topic similarity.\n\n**3rd place ($10,000)**\n**Submission:** [Impact Potential Analysis of Water Use Efficiency ](https://www.kaggle.com/iamleonie/impact-potential-analysis-of-water-use-efficiency)\n**Users:** @iamleonie \n\nIn this Water-focussed Python notebook, @iamleonie utilises distance metrics between two dimensions to create a complex impact potential KPI, weighting water use efficiency by the geospatial context a company operates within. This analysis elevates the role of water stressed geographies to a company’s KPI performance and leverages external datasets from [WRIs Aquaduct tool](https://www.wri.org/aqueduct) .The notebook then uses these metrics brilliantly to rank countries, cities, corporations, and facilities according to both their performance and their potential for improvement. \n\n## Honourable Mentions\n\n**Submission:** [KPIs for Measuring Climate Action and Inequality](https://www.kaggle.com/mannmann2/kpis-for-measuring-climate-action-and-inequality)\n**Users:** @mannmann2 ; @rebeccaverghese \n\nIn this comprehensive Python notebook, @mannmann2 and @rebeccaverghese create an extensive set of KPIs using the CDP dataset to explore a range of different environmental and economic issues.\n\n\n**Submission:**[Tree Canopy Equality](https://www.kaggle.com/adriansarstedt/cdp-tree-canopy-equality-kpi-final-submission)\n**Users:** @adriansarstedt ; @hamishgunasekara \n\nThis Python notebook takes an innovative approach to the challenge by looking at tree canopy equality within cities and how that perpetuates existing imbalances between the rich and the poor.\n\n**Submission:** [Do You Have a Climate Plan?](https://www.kaggle.com/jbrans/do-you-have-a-climate-plan)\n**Users:** @jbrans \n\nIn this Python notebook, @jbrans develops a number of different KPIs and visualises combined KPI scores for every city in the CDP dataset in an interactive map. KPI scores are plotted against “percent of population living in poverty” to intelligently explore the intersection between environmental issues and social/economic issues.\n\n**Submission:** [CDP Microlevel Urban Sustainability Approach](https://www.kaggle.com/reubencpereira/cdp-microlevel-urban-sustainability-approach)\n**Users:** @reubencpereira \n\nIn this Python notebook @reubencpereira utilized radar plots brilliantly to visualise their calculated CDP sustainability index, presenting a well-automated solution that could be scaled to many Cities across the CDP sample.\n\n**Submission:** [You Can’t Manage What You Don’t Measure](https://www.kaggle.com/vlarmet/cdp-you-can-t-manage-what-you-don-t-measure/)\n**Users:** @vlarmet \n\nIn this Python notebook, @vlarmet develops a sustainability ambition KPI and then maps the scores for every city in the CDP dataset. There is also an extensive narrative that discusses the intersection between environmental issues and social issues.\n\n**Well done to all our winners!**\n\nWe hope you enjoyed the competition and look forward to implementing some of your brilliant ideas within CDP. \n\nHave a great holiday period and stay safe.\n\nRegards,\n\n*The CDP Team*\n\n", "1116006": "Congrats! Will we have feedbacks? I would like to have some please to improve my works.", "1115914": "Congratulations to all winners!\n\nTanks a lot to the competition organizers for offering the opportunity to work on such an interesting and impactful subject. It was a great experience for me working with CDP datasets!\n\nThanks also to the participants who shared their notebooks, insights and datasets!", "1115768": "**Congrats to the winners!**\n\n@shabou, @katemelianova, @artvolgin, and @iamleonie the Kaggle Team will reach out to you shortly with instructions on how to claim your prizes.\n\nThere were a lot of great submissions, thank you everybody for making submissions and for marking your work as public! \n\n", "1116637": "Congratulations to all winners!\n\nWe are thankful to the organisers for holding this competition and offering such rich data for the analysis. Great job to the prize winners and all the participants!\n", "1116040": "@callumr22 @paultimothymooney Thank you very much for the recognition. I feel very honored and had a lot of fun with this challenge. Looking forward to the next one! \n\nCongratulations @shabou @katemelianova @artvolgin to your outstanding work! Great job, guys!\n\nAlso, great job to everyone who has participated. This was an interesting challenge given the amount of data and its structure. There were a lot of great notebooks!", "1115920": "", "1115821": "" } }