Utilizing The Power of Machine Learning in Healthcare
summary
Uses the Kaggle Heart Failure Prediction Dataset and a Colab notebook with TensorFlow/Keras to train a 3-layer neural net (input → 16 → 16 → 1) using one‑hot encoding and StandardScaler, 80/20 train/test split and 100 epochs, yielding about 79% accuracy on held‑out data. Code link and dataset link are provided (Colab, Kaggle). Also builds a simple web input form to host the saved model as a deployable predictor; preprocessing and thresholding (0.5) convert outputs to binary labels.
AI-written summary · about
chapters (10)
- 0:00intro
- 0:22preface
- 0:33topics
- 1:10heart failure background
- 2:38current limitations of diagnosis
- 3:09neural networs
- 5:14heart failure prediction dataset (kaggle)
- 6:29demo
- 15:25conclusion
- 15:47thank you
related talks
17mTransforming Medical Records Into a Heart Attack Prediction Model
20mMachine Learning in Cardiology
39mBest Practices for Prototyping ML Models for Healthcare
57mHow To Evaluate and Improve Machine Learning Models
18mMedical AI Beyond Prediction: From Normative Learning to Agentic Healthcare AI
04mBlack Box: How AI will Amplify the Best and Worst of Humanity
34mWhy AI in the Medical Field Both Sucks and Rocks...
46mIntroduction to Machine Learning