Monsoon 2020: Completed #classroom Weight: 35.0

# DA Project LIGHT

Predict the class of the LED bulb

1272
183
83
2423

## 🕵️ Introduction

The way energy is being wasted, soon light bills might be the scariest thing for a lot of us. Not an extreme but a simple temporary solution can be switching to better lights. So, this challenge brings you attributes and asks you to predict the `class of the LED bulb`.

## 💾 Dataset

The database contains various attributes about LED lights which are used almost everywhere. The database classifies the LED's into 10 different types of `classes` from `0 to 9`. All the attributes are nominal types and all have 2 unique values 0 or 1. There are in total 25 attributes out of which `24` are the nominal 0 or 1 types and the last one is the class of the lED light.

For simplification, all the attributes have been stored in the CSV file which has `24` columns, the last column is the `class` and the rest `23` contain the information about the LED.

## 📁 Files

Following files are available in the `resources` section:

• `train.csv` - (`7500` samples) File that should be used for training. It contains the feature representation and the respective classes.
• `test.csv` - (`2500` samples) File that will be used for testing. Unlike the training file it contains only the feature representation of hands and not the classses.

## 🚀 Submission

• Prepare a python file which should produce a csv file with name "submission.csv" containing header as "class" and predicted value as digit between [0…9] representing one of the 10 possible classes to which the LED belongs.
• Your submission should read the train and test data (available as environment variables) and should write 'submission.csv' containing the predictions for the test set.
• Sample submission format available in resources section.                                                                                                               Make your first submission here 🚀 !!

## 🖊 Evaluation Criteria

During evaluation F1 score will be used to test the efficiency of the model where,

$F1 = 2 * \frac{precision*recall}{precision+recall}$

• DA TAs