Constant accuracy with tabular learner Lesson 1

p1x31 · September 1, 2023, 11:44pm

Hello, I’m trying to train binary classifier on continuous tabular data but getting
accuracy of 0.474708 for all epochs. Am I missing something?

Here is the code example and data:

github.com

p1x31/ctci-python/blob/master/yandex/yandex_50289_ml/ml-task-cosmos/e.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>area</th>\n",
       "      <th>perimeter</th>\n",
       "      <th>major_axis</th>\n",
       "      <th>minor_axis</th>\n",
       "      <th>eccentricity</th>\n",
       "      <th>eqdiasq</th>\n",

This file has been truncated. show original