{
  "fetched_at": "2026-10-19T11:00:19Z",
  "source_url": "https://www.kaggle.com/competitions/final-hackathon/data",
  "pages": [
    {
      "name": "data-description",
      "content": "This dataset contains FPS (Frames Per Second) measurements of video games run on various computers. Each row represents the FPS outcome for a specific game executed on a particular computer configuration. The computer's configuration is described by the technical specifications of its CPU and GPU. These specifications include attributes prefixed with \"Cpu\" and \"Gpu,\" respectively, and reflect the factory settings of these components.\n\nThe dataset also characterizes each game by its name, the resolution at which it was displayed, and the quality settings used during the FPS measurement. These attributes are prefixed with \"Game.\"\n\nIn summary, this dataset provides a comprehensive overview of how different computer configurations (CPU and GPU specifications) and game settings (name, resolution, and quality) impact the FPS performance of video games.\n\n## Files\n\n*   **train.csv** - the training set\n*   **test.csv** - the test set\n*   **sample_submission.csv** - a sample submission file in the correct format\n\n## Columns\n\n* `CpuNumberOfCores` - number of physical cores\n* `CpuNumberOfThreads` - number of threads\n* `CpuBaseClock` - base clock in MHz\n* `CpuCacheL1` - total size of level 1 cache in kB\n* `CpuCacheL2` - total size of level 2 cache in kB\n* `CpuCacheL3` - total size of level 3 cache in MB\n* `CpuDieSize` - physical size of the die in square meters\n* `CpuFrequency` - frequency in MHz\n* `CpuMultiplier` - multiplier of CPU\n* `CpuMultiplierUnlocked` - 0=multiplier locked, 1=multiplier unlocked\n* `CpuProcessSize` - used process size in nanometers\n* `CpuSMPNumberOfCpus` - number of symmetric multiprocessors\n* `CpuTDP` - thermal design power in watts\n* `CpuTransistors` - count of transistors in millions\n* `CpuTurboClock` - turbo clock in MHz\n* `GpuBandwidth` - bandwidth in MB/s\n* `GpuBaseClock` - base clock in MHz\n* `GpuBoostClock` - boost clock in MHz\n* `GpuComputeUnits` - number of computing units\n* `GpuDieSize` - physical size of die in square meters\n* `GpuNumberOfExecutionUnits` - number of execution units\n* `GpuFP32Performance` - theoretical Float 32 performance in MFLOP/s\n* `GpuMemoryBus` - width of memory bus in bits\n* `GpuMemorySize` - size of memory in MB\n* `GpuPixelRate` - theoretical pixel rate in MPixel/s\n* `GpuProcessSize` - used process size in nanometers\n* `GpuNumberOfROPs` - number of render output units\n* `GpuNumberOfShadingUnits` - number of shading units\n* `GpuNumberOfTMUs` - number of texture mapping units\n* `GpuTextureRate` - theoretical texture rate in KTexel/s\n* `GpuTransistors` - number of transistors in millions\n* `GpuArchitecture` - architecture code\n* `GpuMemoryType` - memory type\n* `GpuBusInterface` - bus interface\n* `GpuOpenCL` - version of OpenCL\n* `GpuShaderModel` - version of shader model\n* `GpuVulkan` - version of Vulkan\n* `GpuOpenGL` - version of OpenGL\n* `GameName` - name\n* `GameResolution` - resolution\n* `GameSetting` - quality setting\n* `Dataset` - indicator for the source of measurement\n* `FPS` - FPS value",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "IITG.ai is a platform for the AI community of IITG to discuss, learn and collaborate on advancements in the field of AI. We have created a culture of contributing to and improving AI's dynamic, fast-changing world.\n\nHere is a handy link to Kaggle's [competition documentation](https://www.kaggle.com/docs/competitions), which includes, among other things, instructions on [submitting predictions](https://www.kaggle.com/docs/competitions#making-a-submission).\n\nBefore starting make sure you go through the [Rules](https://www.kaggle.com/competitions/final-hackathon/rules) to understand the format of the leaderboard and know about the rules regarding submissions.\n\nAlso, we have provided a [starter notebook](https://www.kaggle.com/himanshusinghal19/ml-ai-week-7-starter-notebook) to get an idea of how to approach this competition.\n\nTo understand how your submissions are scored on the leaderboard, go through the evaluation tab, we have provided details on the metrics used to score your submission.\n\nThe score given to you on the course leaderboard for this hackathon will be dependent on your score on the private leaderboard, we will also require notebook submissions from all participants. The course leaderboard scores will depend significantly on the quality of your work you show, not just your private LB score.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The evaluation metric for this competition is [Mean Squared Error (MSE)](https://en.wikipedia.org/wiki/Mean_squared_error). MSE is a commonly used measure of the differences between predicted and observed values. It is calculated as the average of the squared differences between the prediction and the actual observation. The formula for MSE is given by:\n\n$$\\text{MSE} = \\frac{1}{n} \\sum_{i=1}^{n} (y_i - \\hat{y}_i)^2$$\n\nwhere \n$$y_i \\text{ = actual value}, \\newline\n\\hat{y}_i \\text{ = predicted value, and} \\newline\nn \\text{ = the number of observations.}$$\n\nMSE gives a relatively high weight to large errors, making it particularly useful when large errors are especially undesirable.\n\n## Submission File\nFor each ID in the test set, you must predict a particular value. The file should contain a header and have the following format:\n\n    Index,value\n    2,5.63\n    5,3.14\n    6,7.89\n    etc.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "Rules and guidelines should be strictly followed for the hackathon. Breaking rules may lead to disqualification from the Hackathon.\n- This is an open hackathon where every person participates **individually,** so no team of 2 and above formation is allowed.\n- You'll be provided with a **Training** as well as **Testing** Data Set. You have to Train your model on the training data set and submit the predictions over\nthe test data set on the hackathon's page.\n- There will be 2 Leaderboards:\n  - **Public Leaderboard** : The Public leaderboard will be visible to everyone throughout the time\n  - **Private Leaderboard** : It will be visible once the hackathon ends. Public leaderboards judge your predictions on **60%** of the test data set and private\nleaderboard judges on the rest **40%.** The reason behind the Private leaderboard is to discourage the practice of overfitting the model on the Public\nleaderboard\n- You are allowed to make only **10 submissions in a day**.\n- You all are supposed to **choose two submission** on before the end of competition.\n- All participants of the competition will be asked to submit their notebooks so keep it informative and clean. **Your Notebook submissions will also be evaluated.**\n- A **Starter Notebook** is published to give you an idea about the approach.\n- This **hackathon will be part of your grading** so you are requested to keep your team name as **\"UniqueID_First Name(Example: 2098_John)\"** while\nsubmitting on Kaggle.\n- Everyday, you'll be allowed to make a maximum of only 10 submissions on the website. So, it'd be good if you start earlier. The people who will start\nsoon will be able to do more submissions.\n- You're strictly advised **to not publish your notebooks before the end of the Hackathon**.If found, it'd lead to disqualification of your team.\n- You're **not allowed to discuss** the approach among the participants in the \"Discussion Channel\" before the end of Hackathon.\n\nTLDR\n\n- Don't cheat!\n- Apply yourself!\n- Have Fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "This competition is the Final Hackathon as part of Machine Learning.AI 2024, a primer course on Data Science and Machine Learning, conducted by IITG.ai club.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}