I know that this question is kind of a basic, but I will be glad if some one tells me what will be the target variable (y) in this competition.\nIn the Leaderboard, I see that Laplace Log Likelihood is being used as score to rank the contenders. Please let me know if \"Laplace Log Likelihood.\" is the target variable. \nI would be gratefull if you also point me out for some helpful online resources on the target variable.
", "messages": [ { "id": "945908", "postDate": "07/26/2020 07:56:57", "content": "I know that this question is kind of a basic, but I will be glad if some one tells me what will be the target variable (y) in this competition.\nIn the Leaderboard, I see that Laplace Log Likelihood is being used as score to rank the contenders. Please let me know if \"Laplace Log Likelihood.\" is the target variable. \nI would be gratefull if you also point me out for some helpful online resources on the target variable.
", "rawMarkdown": "I know that this question is kind of a basic, but I will be glad if some one tells me what will be the target variable (y) in this competition.\nIn the Leaderboard, I see that Laplace Log Likelihood is being used as score to rank the contenders. Please let me know if \"Laplace Log Likelihood.\" is the target variable. \nI would be gratefull if you also point me out for some helpful online resources on the target variable.", "votes": null }, { "id": "946660", "postDate": "07/26/2020 17:59:07", "content": "The target variable is the FVC for a given patient/week (for each patient you need to predict every week from -12 to 133).\nLaplace Log Likelihood is the scoring metric; it's a function of the absolute errors of your predictions on the hidden test set and of the confidence levels you give for each of those predictions.
", "rawMarkdown": "The target variable is the FVC for a given patient/week (for each patient you need to predict every week from -12 to 133).\nLaplace Log Likelihood is the scoring metric; it's a function of the absolute errors of your predictions on the hidden test set and of the confidence levels you give for each of those predictions.", "votes": null }, { "id": "947080", "postDate": "07/27/2020 04:21:22", "content": "Hi Mathieu,\nThank you for your answers, it was a great help. Please explain one more thing, How did you obtained this range of -12 to 133 (weeks) any reference or elaboration is highly appreciated. \nCan you also tell me, It is asked to predict the final three FVC measurements as \"You are asked to predict the final three FVC measurements for each patient, .........\" \nMy question is, are we going to use the average of the final three FVC measurements in the Laplace Log Likelihood calculation for the final submission?
", "rawMarkdown": "Hi Mathieu,\nThank you for your answers, it was a great help. Please explain one more thing, How did you obtained this range of -12 to 133 (weeks) any reference or elaboration is highly appreciated. \nCan you also tell me, It is asked to predict the final three FVC measurements as \"You are asked to predict the final three FVC measurements for each patient, .........\" \nMy question is, are we going to use the average of the final three FVC measurements in the Laplace Log Likelihood calculation for the final submission?", "votes": null }, { "id": "947904", "postDate": "07/27/2020 14:46:53", "content": "Thanks both for the discussion. The range of -12 to 133 can be obtained from the sample submission file in the data section.\nYou provide your prediction for each week in this range, and for evaluation, we select the three weeks that correspond to the final three FVC measurements of each patient. Your score is the average of the Laplace Log Likelihood across all patients (average of the final three FVC measeurements). Hope that helps.
", "rawMarkdown": "Thanks both for the discussion. The range of -12 to 133 can be obtained from the sample submission file in the data section.\nYou provide your prediction for each week in this range, and for evaluation, we select the three weeks that correspond to the final three FVC measurements of each patient. Your score is the average of the Laplace Log Likelihood across all patients (average of the final three FVC measeurements). Hope that helps.", "votes": null } ], "comments": [ { "id": 946660, "author_name": "archimedus", "author_url": "", "post_date": "07/26/2020 17:59:07", "content": "The target variable is the FVC for a given patient/week (for each patient you need to predict every week from -12 to 133).\nLaplace Log Likelihood is the scoring metric; it's a function of the absolute errors of your predictions on the hidden test set and of the confidence levels you give for each of those predictions.
", "votes": null, "replies": [ { "id": 947080, "author_name": "tauqeerkhurram", "author_url": "", "post_date": "07/27/2020 04:21:22", "content": "Hi Mathieu,\nThank you for your answers, it was a great help. Please explain one more thing, How did you obtained this range of -12 to 133 (weeks) any reference or elaboration is highly appreciated. \nCan you also tell me, It is asked to predict the final three FVC measurements as \"You are asked to predict the final three FVC measurements for each patient, .........\" \nMy question is, are we going to use the average of the final three FVC measurements in the Laplace Log Likelihood calculation for the final submission?
", "votes": null, "replies": [] }, { "id": 947904, "author_name": "ahmedhshahin", "author_url": "", "post_date": "07/27/2020 14:46:53", "content": "Thanks both for the discussion. The range of -12 to 133 can be obtained from the sample submission file in the data section.\nYou provide your prediction for each week in this range, and for evaluation, we select the three weeks that correspond to the final three FVC measurements of each patient. Your score is the average of the Laplace Log Likelihood across all patients (average of the final three FVC measeurements). Hope that helps.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "945908": "I know that this question is kind of a basic, but I will be glad if some one tells me what will be the target variable (y) in this competition.\nIn the Leaderboard, I see that Laplace Log Likelihood is being used as score to rank the contenders. Please let me know if \"Laplace Log Likelihood.\" is the target variable. \nI would be gratefull if you also point me out for some helpful online resources on the target variable.", "946660": "The target variable is the FVC for a given patient/week (for each patient you need to predict every week from -12 to 133).\nLaplace Log Likelihood is the scoring metric; it's a function of the absolute errors of your predictions on the hidden test set and of the confidence levels you give for each of those predictions.", "947080": "Hi Mathieu,\nThank you for your answers, it was a great help. Please explain one more thing, How did you obtained this range of -12 to 133 (weeks) any reference or elaboration is highly appreciated. \nCan you also tell me, It is asked to predict the final three FVC measurements as \"You are asked to predict the final three FVC measurements for each patient, .........\" \nMy question is, are we going to use the average of the final three FVC measurements in the Laplace Log Likelihood calculation for the final submission?", "947904": "Thanks both for the discussion. The range of -12 to 133 can be obtained from the sample submission file in the data section.\nYou provide your prediction for each week in this range, and for evaluation, we select the three weeks that correspond to the final three FVC measurements of each patient. Your score is the average of the Laplace Log Likelihood across all patients (average of the final three FVC measeurements). Hope that helps." }, "source": "meta" }