I tried to run a simple vgg19 pretrained model for this competition using PyTorch. I got all the prediction probabilities to be a single value.
\n\nIs this due to the heavy imbalance in the data, or something else may be causing this?
\n\nAlso, accuracy was stuck at one point too, while loss was normally decreasing over each epoch.
", "messages": [ { "id": "869009", "postDate": "05/31/2020 16:25:26", "content": "I tried to run a simple vgg19 pretrained model for this competition using PyTorch. I got all the prediction probabilities to be a single value.
\n\nIs this due to the heavy imbalance in the data, or something else may be causing this?
\n\nAlso, accuracy was stuck at one point too, while loss was normally decreasing over each epoch.
", "rawMarkdown": "I tried to run a simple vgg19 pretrained model for this competition using PyTorch. I got all the prediction probabilities to be a single value.\n\nIs this due to the heavy imbalance in the data, or something else may be causing this?\n\nAlso, accuracy was stuck at one point too, while loss was normally decreasing over each epoch.", "votes": null }, { "id": "869697", "postDate": "06/01/2020 08:06:29", "content": "Unbalanced data might seem to be introducing a bias in your model but it cannot be causation for poor accuracy results. \nArtificially introducing balance in your dataset will result in your model not learning comprehensively about the dataset and hence leading to poor predictions. \nA hypothesis could be tested after creating balance and checking out the normalized confusion matrix for your model. I faced the same problem while dealing with my notebook here. \nAfter a bit of research this article explained to me why unbalanced data might not be a reason for my model giving poor results.\nI hope you find it helpful. Also, do check out my first post in my new series named Beginners' Mistakes over here. Thank you :)
", "rawMarkdown": "Unbalanced data might seem to be introducing a bias in your model but it cannot be causation for poor accuracy results. \nArtificially introducing balance in your dataset will result in your model not learning comprehensively about the dataset and hence leading to poor predictions. \nA hypothesis could be tested after creating balance and checking out the normalized confusion matrix for your model. I faced the same problem while dealing with my notebook [here](https://www.kaggle.com/navinmundhra/beginners-mistakes-01-multiclass-lgbm-tuned). \nAfter a bit of research [this](https://matloff.wordpress.com/2015/09/29/unbalanced-data-is-a-problem-no-balanced-data-is-worse/) article explained to me why unbalanced data might not be a reason for my model giving poor results.\nI hope you find it helpful. Also, do check out my first post in my new series named **Beginners' Mistakes** over [here](https://www.kaggle.com/navinmundhra/beginners-mistakes-01-multiclass-lgbm-tuned). Thank you :)", "votes": null }, { "id": "869709", "postDate": "06/01/2020 08:24:15", "content": "Sure, thanks so much, I will have a look at it.
", "rawMarkdown": "Sure, thanks so much, I will have a look at it.", "votes": null } ], "comments": [ { "id": 869697, "author_name": "navinmundhra", "author_url": "", "post_date": "06/01/2020 08:06:29", "content": "Unbalanced data might seem to be introducing a bias in your model but it cannot be causation for poor accuracy results. \nArtificially introducing balance in your dataset will result in your model not learning comprehensively about the dataset and hence leading to poor predictions. \nA hypothesis could be tested after creating balance and checking out the normalized confusion matrix for your model. I faced the same problem while dealing with my notebook here. \nAfter a bit of research this article explained to me why unbalanced data might not be a reason for my model giving poor results.\nI hope you find it helpful. Also, do check out my first post in my new series named Beginners' Mistakes over here. Thank you :)
", "votes": null, "replies": [ { "id": 869709, "author_name": "sarques", "author_url": "", "post_date": "06/01/2020 08:24:15", "content": "Sure, thanks so much, I will have a look at it.
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "869009": "I tried to run a simple vgg19 pretrained model for this competition using PyTorch. I got all the prediction probabilities to be a single value.\n\nIs this due to the heavy imbalance in the data, or something else may be causing this?\n\nAlso, accuracy was stuck at one point too, while loss was normally decreasing over each epoch.", "869697": "Unbalanced data might seem to be introducing a bias in your model but it cannot be causation for poor accuracy results. \nArtificially introducing balance in your dataset will result in your model not learning comprehensively about the dataset and hence leading to poor predictions. \nA hypothesis could be tested after creating balance and checking out the normalized confusion matrix for your model. I faced the same problem while dealing with my notebook [here](https://www.kaggle.com/navinmundhra/beginners-mistakes-01-multiclass-lgbm-tuned). \nAfter a bit of research [this](https://matloff.wordpress.com/2015/09/29/unbalanced-data-is-a-problem-no-balanced-data-is-worse/) article explained to me why unbalanced data might not be a reason for my model giving poor results.\nI hope you find it helpful. Also, do check out my first post in my new series named **Beginners' Mistakes** over [here](https://www.kaggle.com/navinmundhra/beginners-mistakes-01-multiclass-lgbm-tuned). Thank you :)", "869709": "Sure, thanks so much, I will have a look at it." }, "source": "meta" }