{ "fetched_at": "2026-10-19T08:04:27Z", "source_url": "https://www.kaggle.com/competitions/adams-sose23/data", "pages": [ { "name": "data-description", "content": "##**Data partitioning**\nThe data has been randomly partitioned into a training set and a test set. You should use the labeled training set for developing your models and validating their quality. Next, you should use the developed models to predict the property prices for the listings in the unlabeled test set.\n\nTo prevent overfitting, the test set is divided into a public and a private subset. The leaderboard score of your submission is computed on a private part of the test set. Once the competition is finished, you will see a private score of your submissions. The private score determines the winner of the competition. While the competition is running, you will be able to see the public score of your submissions. In this way, you can track your progress over time, when testing different models.\n\n##**File descriptions**\nThe data consists of three .csv files containing different information:\n\n- **train.csv**: training set with listing features, description and prices\n- **test.csv**: test set with listing features and description\n- **reviews.csv**: [optional for master students, mandatory for PhD students] reviews for listings in both training and test set\n\n##**Data fields**\nFiles train.csv and test.csv have the same columns except for the price variable:\n\n- listing_id: unique property identifier\n- price: [target variable] property price per night in US dollars\n- summary: [text] listing summary\n- description: [text] listing description\n- space: [text] space description\n- neighborhood_overview: [text] neighborhood description\n\nAs indicated above, the text data is provided in several columns devoted to different topics. You will need to think whether to merge them into a single text or consider separately in order to improve your solution.\n\nThe remaining columns contain numeric/categorical features with the listing attributes such as location, host information, reviewer scores and more. You would need to perform some data preprocessing and think how to incorporate the tabular data into your modeling pipeline. The column containing the urls to the property images should be dropped.\n\n##**Additional data**\nThe data on property reviews in reviews.csv is optional for master students. It can be used to further refine the quality of the solution, but master students can also choose to ignore it. By contrast, PhD students are required to include the review data in their submissions.\n\n##**Reviews**\nData dictionary for reviews.csv:\n\n- listing_id: unique property identifier\n- review_id: unique review identifier\n- reviewer_id: unique reviewer identifier\n- comments: [text] review\n\nReviews can be matched with train.csv by listing_id. You would need to think how to deal with the following challenges:\n\n- reviews may come in different languages\n- listings have a different number of reviews\n- aggregating signal from multiple reviews", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Description", "content": "\n##**Overview**\n\nThe learning goal of ADAMS is to familiarize you with recent developments in deep learning and to enable you to use the corresponding methodology for solving real-world problems. In accordance with this objective, the evaluation of the module is based on an assignment.\n\nYour submission will consist of a Jupyter notebook that includes all steps that we typically encounter in a machine learning pipeline (explanatory data analysis, data preparation, parameter tuning, etc.). Grading of the notebook will be based on several criteria including:\na) the quality of the exposition (e.g., a healthy mix of prose, code, comments, math. formalism)\nb) the organization of the notebook\nc) the scope of your analysis\nd) whether you have solved the task in a sound and rigorous manner\n\nThe final assignment is hosted as a Kaggle in-class competition. The main purpose of this setting is to give you the possibility to interactively query the quality of your modeling pipeline by submitting the model predictions to the online leaderboard. You may submit up to 5 entries daily to track the quality of your solution on unseen test data. Please remember that the actual accuracy of your solution is not the main factor that will determine your grade.\n\n##**Modeling task**\nAirbnb is an online marketplace for accommodation rental founded in 2008. Properties hosted on the Airbnb platform are owned by private companies and individuals. The hosts determine the rental price themselves and pay a service fee to Airbnb.\n\nIn this assignment, you will predict property prices for Airbnb listings in London. You will be provided with data on more than 55.000 properties containing the listing attributes (e.g., location, number of rooms, property type), texts with the listing description, and rental price per night.\n\nUsing a broad range of NLP techniques, you will need to build an NN-based model that predicts the price of listings in the test set based on both text and tabular data. If you wish to further refine the accuracy of your model, you will have access to additional data containing the listing reviews.\n\nPlease refer to other sections of the competition page to download the data and read more about competition rules, evaluation and data description.\n\n##**Deadline**\nThe submission deadline is the 2th of September 2023", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Evaluation", "content": "1. Evaluation Metric:
\nYour submissions will be evaluated by computing RMSE (Root Mean Squared Error) between the actual property prices and submitted price predictions.

\n\n2. Submission File:
\n\nYou should submit your predictions in `.csv` format. For every `listing_id` in the test set, you must predict the price of the corresponding listing per night. The file should contain a header and have the following format: ``` listing_id, price 0FEMC4VA5U, 100.00 PQ2BYEO0QR, 100.00 3FP6WFHUBN, 100.00 ... ```

\n\n3. Sample Submission:
\n\nA demo notebook that illustrates how to create a submission is available.", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "rules", "content": " - Don't cheat!\n - Apply yourself!\n - Have fun!", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false } ] }