Introduction

This project aims at building a Placement Prediction system using Artificial Neural Networks.
This is my Final Year Project of my Bachelor's in Computer Engineering.

Dataset

The Dataset consists of the qualitative and the quantitative parameters of the students that are necessary for improving the prediction.

Qualitative Parameters	Quantitative Parameters
Aptitude Skills	Coding Skills
Communication Skills	Technical Skills
Core Knowledge	Academic Performance
Presentation Skills	Programming Skills
Puzzle Solving Skills	Projects
English Proficiency	Internships
Management Skills	Training
	Backlogs

Currently the dataset is provided is sample data and not actual data

Implementation

Python (2.7) is used as it has good compatability with Machine Learning by using scikit learn, numpy, matplotlib libraries.
The 3D plot of qualitative parameters, quantitative parameters and the probability of being placed is mapped in the 3D space and then for prediction.
There are multiple methods of plotting the data, two of which are given in this repository.

1. Averaging the values of the parameters and represent it as a 3D plot. Graph is as shown:
   

2. T-Distributed Stochastic Neighbor Embedding (TSNE) is used to visualize Higher Dimensional data into Lower Dimensions. In this method, the parameters are combined by using TSNE method (sklearn.manifold.TSNE) to obtain a 2D plot. Graph is as shown:
   

3. Principal Component Analysis (PCA) is used to visualize Higher Dimensional data into Lower Dimensions. In this method, the parameters are combined by using TSNE method (sklearn.decomposition.PCA) to obtain a 2D plot. Graph is as shown:
   
   The only difference is that the TSNE preserves the distances and it works better when there are huge number of dimensions.

K-Nearest Neighbour is then applied to the data.