Problem with sample output

anandprakash1091971 · June 20, 2020, 6:29am

sample output of the challenge contains 3002 rows which is wrong because test input contains only 1002 rows
how output is more than input

github.com

CodingJoshi/MachineLearningCodingBlocks/blob/master/🏆 Separating Chemicals Challenge.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",

This file has been truncated. show original

this is my code it is giving accuracy of 1% only
even when i submit sampleoutput.csv as it is
it is giving me accuracy of 49%

please check where it is wrong

thanks

prashant_ml · June 20, 2020, 10:26am

hey @anandprakash1091971 ,

just ignore that part , it is just an example to show.

hey buddy your code is absolutely great , but just a very small mistake .
While creating your output.csv , you have kept if score < 0.5 then append 1 to predictions , instead of it you just need append 0 if score <0.5 and 1 if score >0.5.

If you make these required changes then you will be able to achieve around 98 score.

I hope this would have resolved your doubt.
Thank You and Happy Learning .

anandprakash1091971 · June 20, 2020, 11:45am

okay i understand
thanks for your help

Shivam_27 · July 8, 2020, 3:18pm

you didn’t normalise the data still getting right answer. can you please tell me how?

prashant_ml · July 8, 2020, 6:16pm

hey @Shivam_27,
Its not always that you need to normalize your data.
Sometimes your models work better without normalizing your data and the same happens here also. ]
Normalizing is very helpful when the difference between ranges of values is very high.
As then the model lacks some generalizing power.

I hope it helped you.
Thank You.

Shivam_27 · July 8, 2020, 6:20pm

ok , i got it. Thanks