Skip to content

Latest commit

 

History

History
26 lines (12 loc) · 1.18 KB

README.md

File metadata and controls

26 lines (12 loc) · 1.18 KB

Applied Computational Intelligence Project: Milestone 3

Dataset

Source : Erbil Polytechnic University, University of Kurdistan Hewler

Link : Heart Disease Classification

Objective

The objective in this milestone is to do a preliminary test on Feedforward Neural Network (FNN) and Particle Swarm Optimization Feedforward Neural Network (PSOFNN) for health data classification.

Firstly, I applied FNN to perform heart disease data classification. Since the data has been preprocessed, I've only have to run minimal data transformation which are normalization of the features and the target. Three optimizers (SGD, Adam, and RMSprop) have been tested to compare the training and test results.

Subsequently, I applied PSO to train FNN on the same dataset.

Source code

FNN link : How to Use Keras to Solve Classification Problems with a Neural Network

PSOFNN link : Pyswarms: Training a Neural Network

contact: charlene.lim.kh@gmail.com