Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 14

Autism Prediction

using Machine
Learning
SCHEME OF PRESENTAION
 Abstract
 Introduction
 Background
 Importing Libraries and Dataset
 Data Cleaning
 Exploratory Data Analysis
 Feature Engineering
 Model Training
 Model Evaluation
 Conclusion
ABSTRACT
Research Objective
The primary objective of this research is to utilize machine learning techniques to predict
whether an individual has autism based on specific input features

Methodology Overview
 Data Cleaning
 Feature Engineering
 Model Training and Evaluation
INTRODUCTION

What is Autism?
Autism is a neurological condition characterized by challenges in communication, social
interaction, and behavioral patterns

Why Machine Learning For Autism?


The lack of reliable diagnostic procedures for autism necessitates exploration of machine learning
as a potential tool for predictive diagnosis
BACKGROUND

Autism Spectrum Quotient (AQ)


 The AQ is a standardized assessment tool used for diagnosing autism in adults
 Cumulative scores from these levels help determine the likelihood of an individual being on
the autism spectrum

Choice of Python
Python was chosen as the primary programming language due to its extensive
libraries and frameworks tailored for machine learning tasks
IMPORTING LIBRARIES &
DATASET

 Pandas
 NumPy
 Matplotlib/Seaborn
 Scikit-learn
 XGBoost
 Imblearn
DATA CLEANING
Processing Steps
 Handling Null Values
 Outlier Removal
 Standardization

Ensuring Data Integrity


Cleaning and preparing the dataset, we ensure that subsequent analyses and model training are
based on reliable and consistent data inputs
EXPLORATORY DATA ANALYSIS
MODEL TRAINING
Preparation and Division of Data
 Train-Test Split
 Addressing Imbalance
 Normalization

Selection of Algorithms
Exploration and implementation of various machine learning algorithms (e.g., Logistic
Regression, SVC) to identify the best-performing model for autism prediction
MODEL EVALUATION
Performance Metrics
 Accuracy Assessment
 Confusion Matrix

Comparative Analysis
Evaluation of multiple algorithms based on their
performance metrics, highlighting strengths and
limitations in predicting autism spectrum disorder
CONCLUSION
Key Findings
 Predictive Accuracy
 Implications

Future Directions
 Further refinement and validation of models using larger datasets and diverse demographic
samples
 Exploration of interpretability techniques to understand model decisions and enhance clinical
applicability
REFERENCES
1. https://embrace-autism.com/autism-spectrum-quotient/

2. https://ieeexplore.ieee.org/document/8679454
THANK YOU

You might also like