Skip to content
View Laithwm's full-sized avatar
3 followers · 3 following

Block or report Laithwm

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Laithwm/README.md 
$ initializing laithwm ...
> loading data pipelines
> visual modules online
> transforming chaos into clarity

👋 Hey, I'm Laith

I’m a Data Science student at the National College of Ireland, focused on transforming data into stories that inform and inspire.
My work blends analytics, visualization, and narrative design — because data should do more than describe; it should connect.

📍 Dublin • 🎯 Predictive Analytics · Data Visualization · Machine Learning

⚙️ Tech Stack & Tools

Python Jupyter Notebook SQL Pandas NumPy Scikit-learn Tableau Power BI VS Code Git

📂 Featured Projects

📊 Customer Churn Prediction
Built a complete machine learning pipeline to predict telecom customer churn, including extensive data cleaning, feature engineering, and model comparison (Logistic Regression, Decision Tree, Random Forest). Achieved ~86% accuracy and identified key churn drivers such as contract type and payment method.
→ Python · Scikit-learn · Pandas · Seaborn · EDA · Classification

🚦 Dublin Traffic Volume Prediction
Applied regression models on Smart Dublin SCATS data (500k+ rows) to predict hourly vehicle volume. Implemented preprocessing pipelines, RFECV feature selection, and 10-fold cross-validation. Gradient Boosting (R²=0.46) achieved the best results, revealing strong temporal and spatial traffic patterns.
→ Python · Scikit-learn · Feature Engineering · Regression · Smart City Data

📈 AAPL Stock Price Forecasting
Forecasted Apple’s closing stock prices using ARIMA, SARIMA, and LSTM models. Compared statistical vs deep learning approaches on real-world financial data, where LSTM reduced RMSE by 2.5× compared to ARIMA.
→ Python · TensorFlow/Keras · Statsmodels · Time Series Forecasting · Deep Learning

🧭 Beyond the Data

Outside of data, I’m always looking for stories and systems — whether that’s through content creation, playing pool, or travelling.
Each of these keeps me curious and grounded:

🎥 Content Creation – I enjoy sharing what I learn and seeing how ideas connect when expressed visually.
🎱 Playing Pool – It sharpens focus, patience, and pattern recognition — skills that overlap surprisingly well with analytics.
🌍 Travelling – Experiencing new places challenges perspective and teaches adaptability, both in life and in problem-solving.

Precision in the data. Personality in the story.

📫 Contact

Email: laith.mactari@gmail.com
LinkedIn: linkedin.com/in/laithwm

Popular repositories Loading

  1. Customer-Churn-Prediction Customer-Churn-Prediction Public

    Machine learning project predicting telecom customer churn using EDA, feature engineering, and classification models.

    Jupyter Notebook

  2. Handwritten-Digit-Recognition Handwritten-Digit-Recognition Public

    Deep learning model for handwritten digit classification with a CNN achieving 99.33% test accuracy on MNIST.

    Jupyter Notebook

  3. What-Drives-A-Movies-Success What-Drives-A-Movies-Success Public

    Data visualization project exploring the key factors influencing movie success using IMDb data (1986–2016). Built in Tableau with data preprocessing in Python.

  4. Laithwm Laithwm Public

  5. Dublin-Traffic-ML Dublin-Traffic-ML Public

    Machine learning project predicting hourly traffic volume across Dublin using SCATS sensor data — includes preprocessing, feature engineering, and regression modelling with Gradient Boosting and Ra…

    Jupyter Notebook

  6. AAPL-Stock-Price-Forecasting AAPL-Stock-Price-Forecasting Public

    Time series forecasting of Apple (AAPL) stock prices using ARIMA, SARIMA, and LSTM models — comparing classical and deep learning approaches.

    Jupyter Notebook