{
  "id": 465593,
  "title": "50th Place Solution (Simple Solution) ",
  "url": "/competitions/UBC-OCEAN/discussion/465593",
  "author_name": "Mohammad Essam",
  "post_date": "2024-01-04T19:49:50.920000",
  "votes": 10,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I am happy to achieve my first solo silver competition medal. It was a very good experience to deal with a medical problem like this. </p>\n<h2>My Solution</h2>\n<p>I utilized the public solution provided <a href=\"https://www.kaggle.com/code/jirkaborovec/cancer-subtype-tiles-masks-w-lightning-timm?scriptVersionId=153762364\" target=\"_blank\">here</a> with two modifications : <br>\n1- In training, I used Maxvit Base (achieved 0.48 in both public and private testing)<br>\n2- In Inference, I applied a Softmax threshold technique such that : </p>\n<pre><code>\n max_mean_prob &lt; :\n    row[] =   \n:\n    lb = np.argmax(mean_probs)\n    row[] = labels[lb]\n</code></pre>\n<p>The inference notebook is <a href=\"https://www.kaggle.com/code/mohammedessam97/top50-v3-cancer-subtype-inference?scriptVersionId=154793672\" target=\"_blank\">here</a></p>",
  "messages": [
    {
      "id": 2587521,
      "postDate": "2024-01-04T19:49:50.920Z",
      "content": "<p>I am happy to achieve my first solo silver competition medal. It was a very good experience to deal with a medical problem like this. </p>\n<h2>My Solution</h2>\n<p>I utilized the public solution provided <a href=\"https://www.kaggle.com/code/jirkaborovec/cancer-subtype-tiles-masks-w-lightning-timm?scriptVersionId=153762364\" target=\"_blank\">here</a> with two modifications : <br>\n1- In training, I used Maxvit Base (achieved 0.48 in both public and private testing)<br>\n2- In Inference, I applied a Softmax threshold technique such that : </p>\n<pre><code>\n max_mean_prob &lt; :\n    row[] =   \n:\n    lb = np.argmax(mean_probs)\n    row[] = labels[lb]\n</code></pre>\n<p>The inference notebook is <a href=\"https://www.kaggle.com/code/mohammedessam97/top50-v3-cancer-subtype-inference?scriptVersionId=154793672\" target=\"_blank\">here</a></p>",
      "rawMarkdown": "I am happy to achieve my first solo silver competition medal. It was a very good experience to deal with a medical problem like this. \n\n## My Solution \nI utilized the public solution provided [here](https://www.kaggle.com/code/jirkaborovec/cancer-subtype-tiles-masks-w-lightning-timm?scriptVersionId=153762364) with two modifications : \n1- In training, I used Maxvit Base (achieved 0.48 in both public and private testing)\n2- In Inference, I applied a Softmax threshold technique such that : \n\n```python\n# Decide if it's an outlier\nif max_mean_prob < 0.3:\n    row['label'] = 'Other'  # Assign to outlier\nelse:\n    lb = np.argmax(mean_probs)\n    row['label'] = labels[lb]\n```\nThe inference notebook is [here](https://www.kaggle.com/code/mohammedessam97/top50-v3-cancer-subtype-inference?scriptVersionId=154793672)",
      "votes": 10
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2587521": "I am happy to achieve my first solo silver competition medal. It was a very good experience to deal with a medical problem like this. \n\n## My Solution \nI utilized the public solution provided [here](https://www.kaggle.com/code/jirkaborovec/cancer-subtype-tiles-masks-w-lightning-timm?scriptVersionId=153762364) with two modifications : \n1- In training, I used Maxvit Base (achieved 0.48 in both public and private testing)\n2- In Inference, I applied a Softmax threshold technique such that : \n\n```python\n# Decide if it's an outlier\nif max_mean_prob < 0.3:\n    row['label'] = 'Other'  # Assign to outlier\nelse:\n    lb = np.argmax(mean_probs)\n    row['label'] = labels[lb]\n```\nThe inference notebook is [here](https://www.kaggle.com/code/mohammedessam97/top50-v3-cancer-subtype-inference?scriptVersionId=154793672)"
  }
}