Skip to content

Gloireski/symptom_reader

Repository files navigation

symptom_reader

Introduction

The Project

In today's digital world, we often rely on the internet to explain our symptoms and provide insights into potential health issues. Symptom_Reader aims to bridge this gap by offering reliable diagnoses and holistic treatment options, including OTC medications, in a single click. While no web application can replace a professional medical diagnosis, Symptom_Reader assists users in understanding their symptoms and determining preliminary actions before consulting a doctor.

symptom_reader is a cutting-edge web application designed to deliver accurate medical diagnoses based on user-reported symptoms. It offers comprehensive home remedies and over-the-counter medication recommendations for various medical conditions. Inspired by the need for accessible and reliable health information, Symptom_reader seamlessly integrates holistic and medical treatment options. Beyond identifying illnesses, Symptom_reader empowers users with the knowledge to effectively manage and treat their symptoms. This is the core purpose of Symptom_Reader. Take your Health seriously.

The Context

This project serves as our Portfolio Project, marking the conclusion of our Foundations Year at Holberton School. We had the freedom to select our teammates and choose the focus of our work, provided we delivered a functioning program by the end of the three-week development period.

The Team

We are three software Engineers who are passionate about coding and our diverse careers and cultures from three different African countries.

  • Amanda Mabunda is a South African citizen, graduated from Biotechnology, a software engineer and founder of Symptom_Reader @Amatas_aura

  • Belem Gloire BEKOUTOU is a Chadian citizen living in Morocco, an IT systems engineering graduate with strong software implementation process experience and co-Founder of Symptom_Reader @GloireBelem

  • Ado Abdullateef is a Nigerian citizen, I am a remote sensing and GIS graduate, geospatial analyst, software engineer, and co-founder of Symptom_Reader. @adoabdullateef

follow us on Twitter for more.

Blog Posts

After the development phase, we each wrote a blog post to reflect on the development journey of Symptom_Reader

Tutorial

Symtom_Reader

Here is a preview of our main feature, a results page with the user's diagnosis and recommendations

Results page

Here is a simple flow for the user experience on Symptom_Reader

Create a simple flow diagram

Architecture

Overview

Our web application predominantly employs HTML, CSS, and JavaScript for front-end development. We prioritized simplicity and user-friendliness, focusing on crafting intuitive interfaces without heavy reliance on frameworks. Bootstrap streamlined tasks such as user authentication. Python Flask handles backend operations, including routing and data processing. Machine learning, based on Kaggle disease prediction datasets, aids in disease prediction. Our approach emphasizes a balanced blend of technology to ensure a seamless and effective user experience.

Create a simple flow diagram

Final model training

We used training and testing datasets from kaggle, after doing some earlier training on 80% of the data and make sure our models worked well using confusion matrix plots we started final training on the whole dataset. To achieve an accurate and robust model we combined three machine learning algorithms: Support Vector Classifier, Gaussian Naive Bayes Classifier and Random Forest Classifier. We used label encoder to convert the prognosis column to numerical variable which is more suitable for model training. We finally put the column in dictionary for make sure we have the right name in the inputs.

Reading input data and predict

finally, we’ve created a function predict_disease that takes as a parameter a string of metrics separated by a comma, the metrics name present will be converted to 1 and 0 for the final prediction is done accordingly by combining our three algorithms

About

this is our Alx foundation porfolio project repo

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published