Project

Expanded Diseased Leaf Detection Project on STM32 Embedded Machine Learning System (2022) - Python, C

• Proposed an effective apporoach to fit the dataset and enhance model’s accuracy under 1MB memory constraint.

• Optimized the weights and biases when deploying the model on embedded system.

• Implemented successfully the 2D CNNs with high accuracy.

Diseased Leaf Detection and Classification (2022) - Python

• Collected, extracted and analyzed data using Pandas, NumPy libraries.

• Created and developed an Ensemble Learning of EfficientNetB7 and Exception model.

• Enhanced the model accuracy by taking the average values to achieve 96%.

Handwritten Digits Recognition (2021) - Python, C - STM32 H743ZI2

• Developed a CNN model to classify handwritten digits in Python.

• Designed and optimized the Convolutional layer, Max pooling layer and Dense layer in C.

• Deployed the weights and biases of the pre-trained model into the Microcontroller STM32 H743ZI2 board.

Spam Email Classification (2021) - Python

• Built a model to classify an email to be spam or non-spam based on Naiive Bayes Theorem.

• Optimized the model by using Scikit-learn library

• Accuracy: 97%

House Prices Prediction (2021) - Python

• Extracted, analyzed and visualized the dataset.

• Developed a data transformation to optimize the Linear Regression Model.

• Built a valuation tool for the prices prediction.

Fire Warning Device (2016) - C

• Developed an electronic device using Micro-controller 8051

• The device is a combination of measuring temperature, humidity, gas, current and displays on a LCD soldiered on the circuit.

• The device will alarm through a speaker when the temperature or gas concentration increases rapidly and passes the threshold.