Gerald’s Portfolio

Hi there, I’m Gerald! 👋🏻 Welcome to my personal portfolio page. 🙏🏻
Below are the projects I’ve done:

Project #4
House Price Prediction in Bengaluru (Python)

Conducted outlier handling, missing value imputation, implemented one-hot encoding for categorical variables, and managed data type handling in the preprocessing stage.
Performed and compared three regression algorithms: Linear Regression, Lasso, and Decision Tree Regressor.
Calculated scores and searched for the best parameters for each algorithm using GridSearchCV. The result indicated that Linear Regression was the most suitable for this case with a score of 81.9%

***

Project #2
Image Classification of Rock-paper-scissors Hand-gesture Pictures (Python)

Built a CNN model using TensorFlow and Keras libraries to classify hand gesture photos as either forming a rock, paper, or scissors
Achieved accuracy and validation accuracy results exceeding 96%
Utilized image augmentation techniques and callbacks, resulting in accuracy and validation accuracy surpassing 96%
Received a perfect score of 5 out of 5 stars from the reviewer. This is the final project for the machine learning course from Dicoding Indonesia.

***

Project #1
Visualization and Analysis of an Online Retail Invoices Dataset (Python)

Conducted exploratory data analysis on 541,909 sales records with Pandas library
Performed data preprocessing to address missing values, non-positive sales figures, ambiguity in product descriptions, and data type inconsistencies
Performed visualization of monthly and daily sales trends, top selling products, and top loyal customers
Utilized Matplotlib and Seaborn libraries for visualizing monthly and daily sales trends, top-selling products, and top loyal customers
Extracted insights and provided recommendations from the visualization results

***