Autism Spectrum Disorder, ASD, has witnessed several developments since its initial study in the early 1900s. ASD is characterized by difficulty in communication and socializing and might sometimes include learning disabilities. Sequel to the recent developments in the study of Autism and the adoption of Machine Learning to proferring solutions in healthcare, Machine Learning was employed in this project to help predict the diagnosis of Autism across all age categories: Toddlers, Children, Adolescents and Adults. The dataset has its source from Dr. Fadi Thabtah from Manukau Institute of Technology. The project adopts the AQ-10 and Q-CHAT-10 surveys for diagnosis. The machine learning model for toddlers was deployed into a Web Application for the purpose of reproduction and presentation and can be found here.
- notebook_file.ipynb
This contains the codes for the analysis, modelling and evaluation in Jupyter Notebook
- model.pickle
This is the best model for predicting diagnosis in toddlers data, saved as a pickle file
- toddler_app.py
This is the python file for deploying the toddler diagnosis model as a web application
- Link to Web Application
The link to the Web application can be found here : https://awojidetola-autism-prediction-toddler-app-dng4yi.streamlitapp.com/