Skip to content

This project analyzes fictional recruiting data through two main approaches - prediction and explanation. A multilayer perceptron is used for prediction and logistic regression for explanation.

Notifications You must be signed in to change notification settings

gzlupko/Recruiting_Prediction_Neural_Network

Repository files navigation

Overview

In this project I analyze fictitious recruiting data for two separate tasks - prediction and explanation. In both tasks, the models that are used aim to identify which candidates will be hired based on a series of input variables. In the prediction tasks, a multilayer perceptron neural network model is trained with the goal of optimizing prediction performance. While the neural network predicts accurately 'off the shelf', the second task in this project is to use an alternative model that aids interpretation of the factors that influence hiring decisions for this company. For the explanation task, a multiple logistic regression model is trained to predict hiring decisions but also to be able to enable a close examination of the relationships between the input variables on the outcome.

About

This project analyzes fictional recruiting data through two main approaches - prediction and explanation. A multilayer perceptron is used for prediction and logistic regression for explanation.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published