{
  "id": 543959,
  "title": "Variable Importance Heatmap with Automated Machine Learning",
  "url": "/competitions/child-mind-institute-problematic-internet-use/discussion/543959",
  "author_name": "Taimour Nazar",
  "post_date": "2024-11-02T13:25:16.909000",
  "votes": 0,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hello</p>\n<p>Reference Notebook Version 42: <a href=\"https://www.kaggle.com/code/taimour/automl-h2o-feature-engineering-piu\" target=\"_blank\">🌐 AutoML H2O - ⚙️ Feature Engineering - 📱 PIU</a> </p>\n<p>Below is the variable importance heatmap which I have genereted with AutoML H2O. Different models are run and after that this heatmap is generated. Y-axis shows features and X-axis shows models. Overall heatmap shows how much a feature was considered important by different models. </p>\n<p>My question is that can we select features of our choice for the heatmap with AutoML H2O or will the H2O algorithm always select the displayed features by itself? </p>\n<p><em>Note: Heatmaps can vary between different versions of notebook due to different models selected and due to new features added after combining different features with mathematical operations.</em></p>\n<p><img src=\"https://www.kaggleusercontent.com/kf/204708891/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..Y56riTCEYL5-bjkkJs8DLQ.JzEnsWDT7veNvx9UiQQKFvEJDmq0_pqcow7Ek2dfm6CVBkx882lyc8jyWDtBcCYG2kPZ1ySUU1hEbtZExH7-_qztGZ87L_ry1kuhzM7sn4g0M45TvIcJO3DGZ8NW949c-iGM2axeRIBhoaGLE4-Gi5-2kUBJVyM2AyZtzKK7Kt0Cdy5MZkQSF7xKlpsLjnuTeof44kTCZCUdwUcT86hGpe73b9HedaEa4DOpvs8JZsQW8xsTeVW0GN2Alj0do-SVYYBlUKclG76_GFumF1P6EmmdbIRQfPgpBs5ofyQDpHt4btojZmSRgJzZwqjAUbeLNpNKPMygV4iGvURfY0tmk14hqF4yvBKVbnWFl5eR4Oa_T1yL3ybf3IP4BYQcRGSzXZg79z6DwsbsrRBwZPp07Vuw6vJRPgcvU0MB2k6446q6dB9CHeRXVpin2tmA9qngepry-J2k9YGn9P8N089u_k-egYri7WUQdxSsRLOhettOHV6_bV_Fa_Jmde3r2i0s0kUhN5WBLpL6UjFzfs7cUJyQNulKx4kWB_LZj2VAGMQrxWp3eaJCdsmVp9v5IylWbzQ91Qg7f9TbYIaV38cbf3PuU-ofF0hFO2LBD5aAC4oKjKdpSMKm0mawEphEdEXDciWr7ubxtFOPFTggqG-zsw.VIhP5FNfYzOKiWNC2skP_w/__results___files/__results___32_1.png\" alt=\"\"></p>",
  "messages": [
    {
      "id": 3034685,
      "postDate": "2024-11-02T13:25:16.910Z",
      "content": "<p>Hello</p>\n<p>Reference Notebook Version 42: <a href=\"https://www.kaggle.com/code/taimour/automl-h2o-feature-engineering-piu\" target=\"_blank\">🌐 AutoML H2O - ⚙️ Feature Engineering - 📱 PIU</a> </p>\n<p>Below is the variable importance heatmap which I have genereted with AutoML H2O. Different models are run and after that this heatmap is generated. Y-axis shows features and X-axis shows models. Overall heatmap shows how much a feature was considered important by different models. </p>\n<p>My question is that can we select features of our choice for the heatmap with AutoML H2O or will the H2O algorithm always select the displayed features by itself? </p>\n<p><em>Note: Heatmaps can vary between different versions of notebook due to different models selected and due to new features added after combining different features with mathematical operations.</em></p>\n<p><img src=\"https://www.kaggleusercontent.com/kf/204708891/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..Y56riTCEYL5-bjkkJs8DLQ.JzEnsWDT7veNvx9UiQQKFvEJDmq0_pqcow7Ek2dfm6CVBkx882lyc8jyWDtBcCYG2kPZ1ySUU1hEbtZExH7-_qztGZ87L_ry1kuhzM7sn4g0M45TvIcJO3DGZ8NW949c-iGM2axeRIBhoaGLE4-Gi5-2kUBJVyM2AyZtzKK7Kt0Cdy5MZkQSF7xKlpsLjnuTeof44kTCZCUdwUcT86hGpe73b9HedaEa4DOpvs8JZsQW8xsTeVW0GN2Alj0do-SVYYBlUKclG76_GFumF1P6EmmdbIRQfPgpBs5ofyQDpHt4btojZmSRgJzZwqjAUbeLNpNKPMygV4iGvURfY0tmk14hqF4yvBKVbnWFl5eR4Oa_T1yL3ybf3IP4BYQcRGSzXZg79z6DwsbsrRBwZPp07Vuw6vJRPgcvU0MB2k6446q6dB9CHeRXVpin2tmA9qngepry-J2k9YGn9P8N089u_k-egYri7WUQdxSsRLOhettOHV6_bV_Fa_Jmde3r2i0s0kUhN5WBLpL6UjFzfs7cUJyQNulKx4kWB_LZj2VAGMQrxWp3eaJCdsmVp9v5IylWbzQ91Qg7f9TbYIaV38cbf3PuU-ofF0hFO2LBD5aAC4oKjKdpSMKm0mawEphEdEXDciWr7ubxtFOPFTggqG-zsw.VIhP5FNfYzOKiWNC2skP_w/__results___files/__results___32_1.png\" alt=\"\"></p>",
      "rawMarkdown": "Hello\n\nReference Notebook Version 42: [🌐 AutoML H2O - ⚙️ Feature Engineering - 📱 PIU](https://www.kaggle.com/code/taimour/automl-h2o-feature-engineering-piu) \n\nBelow is the variable importance heatmap which I have genereted with AutoML H2O. Different models are run and after that this heatmap is generated. Y-axis shows features and X-axis shows models. Overall heatmap shows how much a feature was considered important by different models. \n\nMy question is that can we select features of our choice for the heatmap with AutoML H2O or will the H2O algorithm always select the displayed features by itself? \n\n*Note: Heatmaps can vary between different versions of notebook due to different models selected and due to new features added after combining different features with mathematical operations.*\n\n![](https://www.kaggleusercontent.com/kf/204708891/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..Y56riTCEYL5-bjkkJs8DLQ.JzEnsWDT7veNvx9UiQQKFvEJDmq0_pqcow7Ek2dfm6CVBkx882lyc8jyWDtBcCYG2kPZ1ySUU1hEbtZExH7-_qztGZ87L_ry1kuhzM7sn4g0M45TvIcJO3DGZ8NW949c-iGM2axeRIBhoaGLE4-Gi5-2kUBJVyM2AyZtzKK7Kt0Cdy5MZkQSF7xKlpsLjnuTeof44kTCZCUdwUcT86hGpe73b9HedaEa4DOpvs8JZsQW8xsTeVW0GN2Alj0do-SVYYBlUKclG76_GFumF1P6EmmdbIRQfPgpBs5ofyQDpHt4btojZmSRgJzZwqjAUbeLNpNKPMygV4iGvURfY0tmk14hqF4yvBKVbnWFl5eR4Oa_T1yL3ybf3IP4BYQcRGSzXZg79z6DwsbsrRBwZPp07Vuw6vJRPgcvU0MB2k6446q6dB9CHeRXVpin2tmA9qngepry-J2k9YGn9P8N089u_k-egYri7WUQdxSsRLOhettOHV6_bV_Fa_Jmde3r2i0s0kUhN5WBLpL6UjFzfs7cUJyQNulKx4kWB_LZj2VAGMQrxWp3eaJCdsmVp9v5IylWbzQ91Qg7f9TbYIaV38cbf3PuU-ofF0hFO2LBD5aAC4oKjKdpSMKm0mawEphEdEXDciWr7ubxtFOPFTggqG-zsw.VIhP5FNfYzOKiWNC2skP_w/__results___files/__results___32_1.png)\n"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "3034685": "Hello\n\nReference Notebook Version 42: [🌐 AutoML H2O - ⚙️ Feature Engineering - 📱 PIU](https://www.kaggle.com/code/taimour/automl-h2o-feature-engineering-piu) \n\nBelow is the variable importance heatmap which I have genereted with AutoML H2O. Different models are run and after that this heatmap is generated. Y-axis shows features and X-axis shows models. Overall heatmap shows how much a feature was considered important by different models. \n\nMy question is that can we select features of our choice for the heatmap with AutoML H2O or will the H2O algorithm always select the displayed features by itself? \n\n*Note: Heatmaps can vary between different versions of notebook due to different models selected and due to new features added after combining different features with mathematical operations.*\n\n![](https://www.kaggleusercontent.com/kf/204708891/eyJhbGciOiJkaXIiLCJlbmMiOiJBMTI4Q0JDLUhTMjU2In0..Y56riTCEYL5-bjkkJs8DLQ.JzEnsWDT7veNvx9UiQQKFvEJDmq0_pqcow7Ek2dfm6CVBkx882lyc8jyWDtBcCYG2kPZ1ySUU1hEbtZExH7-_qztGZ87L_ry1kuhzM7sn4g0M45TvIcJO3DGZ8NW949c-iGM2axeRIBhoaGLE4-Gi5-2kUBJVyM2AyZtzKK7Kt0Cdy5MZkQSF7xKlpsLjnuTeof44kTCZCUdwUcT86hGpe73b9HedaEa4DOpvs8JZsQW8xsTeVW0GN2Alj0do-SVYYBlUKclG76_GFumF1P6EmmdbIRQfPgpBs5ofyQDpHt4btojZmSRgJzZwqjAUbeLNpNKPMygV4iGvURfY0tmk14hqF4yvBKVbnWFl5eR4Oa_T1yL3ybf3IP4BYQcRGSzXZg79z6DwsbsrRBwZPp07Vuw6vJRPgcvU0MB2k6446q6dB9CHeRXVpin2tmA9qngepry-J2k9YGn9P8N089u_k-egYri7WUQdxSsRLOhettOHV6_bV_Fa_Jmde3r2i0s0kUhN5WBLpL6UjFzfs7cUJyQNulKx4kWB_LZj2VAGMQrxWp3eaJCdsmVp9v5IylWbzQ91Qg7f9TbYIaV38cbf3PuU-ofF0hFO2LBD5aAC4oKjKdpSMKm0mawEphEdEXDciWr7ubxtFOPFTggqG-zsw.VIhP5FNfYzOKiWNC2skP_w/__results___files/__results___32_1.png)\n"
  }
}