I'm a Computer Science Engineering student who loves taking an idea from concept → model → application → working prototype.
class AlwinJohn:
role = "CSE Student • AI/ML Developer"
focus = ["Machine Learning", "Computer Vision", "Accessibility Tech"]
mission = "Build software that actually helps people"
status = "Always learning, always shipping 🚀"|
🤖 Artificial Intelligence
|
💻 Software
|
What it does: A healthcare machine-learning system that analyzes patient and lifestyle features to estimate a person's heart attack risk. Multiple classifiers were trained and compared, then combined into a stacked ensemble (SVM, Random Forest, KNN) that beat every individual model.
|
📂 Data |
🧪 Models |
🏆 Result |
📈 Full model comparison (click to expand)
| Model | Accuracy | |
|---|---|---|
| Decision Tree | 79.35% | ▰▰▰▰▰▰▰▱▱▱ |
| Naive Bayes | 83.70% | ▰▰▰▰▰▰▰▰▱▱ |
| Logistic Regression | 84.24% | ▰▰▰▰▰▰▰▰▱▱ |
| KNN | 84.78% | ▰▰▰▰▰▰▰▰▱▱ |
| SVM | 86.96% | ▰▰▰▰▰▰▰▰▰▱ |
| Random Forest | 88.04% | ▰▰▰▰▰▰▰▰▰▱ |
| Voting Ensemble | 88.59% | ▰▰▰▰▰▰▰▰▰▱ |
| Stacking Ensemble | 89.13% | ▰▰▰▰▰▰▰▰▰▰ 🏆 |
What it does: A research-oriented machine-learning project that explores whether postpartum depression can be detected early from socio-demographic and psychosocial survey data. The goal is to turn survey responses into a reliable risk signal that could help flag mothers who may need support.
|
🧹 Clean |
🛠️ Engineer |
🤖 Classify |
🚦 Predict |
What it does: A voice-based computer interaction system designed for people with motor disabilities. Speech is captured and converted with Google Speech-to-Text, parsed into commands, and then used to drive the mouse, keyboard and applications, so a computer can be used completely hands-free.
|
🎤 Listen |
📝 Transcribe |
🧩 Parse |
🖱️ Act |
