Analysis of ADHD among Students

CSEF · 2019 Computational Systems & Analysis (Senior Division Only)

Yuansong Wang · Stevenson (Robert Louis) High School, Pebble Beach, Monterey County

Overview

Objectives Coded for Logistic Regression and Artificial Neuron Network(package) in R language and tested the ADHD model accuracy with ROC. Methods Laptop computer with R language and packages. Analyze the data of 1752 school students from National Health Interview Survey with Logistic Regression model and Artificial Neuron Network and checked the model accuracy with ROC. Results With coefficients' exact numbers of different factors, I can use the logistic regression model to predict a person's probability of getting ADHD, and checked the results with ROC, getting 66 and 69 percent accuracy for logistic regression and Artificial Neuron Network. Conclusions I coded for logistic regression model and Artificial Neuron Network to test the factors of ADHD, and checked the model accuracy with positive distribution and Receiving Operation Characteristics. With my code I can analyze large amount of data to create a predicting model for the issues that have multiple factors.

Summary statement

I coded to analyze data with logistic regression model and Artificial Neuron Network and checked the accuracy of the two models with ROC.

Help received

I programmed the logistic regression myself after an internet search on techniques, and the artificial neuron network and ROC are the packages pf R language tools.