The problem HEART ATTACK CHANCE PREDICTOR solves

Heart attacks can prove to be fatal if not monitored beforehand. A lot of people die due to sudden attacks, which, if predicted earlier, can reduce the risks for the same. However, regularly visiting a doctor is not always feasible for people specially of the older generation. This is the problem we aimed to solved using Machine Learning model that can take day to day inputs and predict your heart health on a regular basis so that you can have a happy and healthy heart.
This will also help in the following ways:
• Help the medical field in developing modern AI based systems to cure diseases.
• Alert the person beforehand about risks and chances of heart attack in the future.
• Use past records and symptoms to generalize the conditions for a particular disease.
• Analyze a person’s heart condition regularly using ML.

Challenges we ran into

We faced the following problems:

1. While training the model and later checking the accuracy of model that we built using Neural Networks in the beginning we observed that it was showing very less accuracy and taking a lot of time to execute. So, we decided to built this model using Logistic Regression algorithm over this dataset and then observed that using this model we have gained a higher accuracy value and precision along with a better time complexity.
2. While taking customized inputs we were facing "Attribute Error" and "shape error" while doing it with Neural Networks, which was then corrected by using Logistic Regression.