Real-Time E-commerce Demand Forecasting
Designed a streaming data pipeline and compared ARIMA, XGBoost, and LSTM for real-time e-commerce demand forecasting, using more than 44,000 transaction records.
MSc thesis · GCTUI'm Francis Amoah Nuako, an MSc Big Data & IoT candidate and software developer based in Accra, Ghana. I work across data analytics, machine learning, software systems, and connected technologies.
01 / ABOUT
My work sits where research questions meet working systems. I enjoy turning complex data into useful insights and building the infrastructure that makes those insights accessible.
My background combines information technology, big data, and healthcare. That mix shapes how I approach problems, from forecasting demand to developing software and working with connected devices.
02 / SELECTED WORK
Thesis research, coursework, independent projects, and professional engineering. More experiments live on GitHub ↗.
Designed a streaming data pipeline and compared ARIMA, XGBoost, and LSTM for real-time e-commerce demand forecasting, using more than 44,000 transaction records.
MSc thesis · GCTUExplored product recommendations and advertising-engagement prediction using machine learning and synthetic customer interaction data.
View repositoryDeveloped an Arduino-based monitoring system to collect environmental sensor readings and visualize changes in air quality.
MSc coursework · GCTULeading development of a listings platform, including structured category data, REST APIs, filtering, and cloud-based deployments.
Co-founder & Lead DeveloperCleaned a public telecommunications extract and measured weekly revenue, customer attributes, and device coverage. The figures describe that historical dataset, not a live network.
View projectCleaned a synthetic hospital admissions file and measured stay length, billing, and a prolonged-stay classifier. The records were generated with Faker and are not real patient data.
View projectCompared Random Forest, Decision Tree, Naive Bayes, and SVM on the UCI Cleveland heart-disease dataset, using leakage-safe preprocessing and one held-out test split. An educational evaluation, not a diagnostic tool.
View projectInvestigated data leakage and class imbalance in payment fraud classification. Compared machine learning models and evaluated how misleading predictors affected performance.
View projectAnalyzed clinical survival data using Kaplan-Meier estimation, log-rank testing, and Cox regression to examine patient survival patterns and associated risk factors.
View projectAnalyzed 2.8 million network flows from CICIDS2017 and compared machine learning models for network attack classification, with additional evaluation of rare and previously unseen attack categories.
View projectInvestigated the vulnerability of image classification models to adversarial attacks using FGSM and PGD, and evaluated how adversarial training affects classification accuracy and robustness.
View projectCards are labeled as thesis research, coursework, an independent project, or professional engineering. Public repositories are linked; unpublished results are not included.
03 / EXPERIENCE
I combine hands-on engineering with research, data analysis, and technical communication.
CITSET Lab · Ghana Communication Technology University
Contributed to institutional IT research through data collection, analysis, cloud-based data workflows, and technical reporting.
Pepapa Ghana Classifieds
Build and maintain backend systems, APIs, database structures, and cloud infrastructure for a classifieds platform.
Digitrust Technologies
Developed business and point-of-sale systems, with work spanning databases, transaction processing, and offline synchronization.
04 / TOOLKIT
Python · SQL · PHP · JavaScript · TypeScript
Pandas · NumPy · Scikit-learn · Streamlit
PostgreSQL · MySQL · MongoDB · Apache Kafka
Laravel · Node.js · Docker · Git · Nginx · Hetzner
05 / BEYOND CODE
I create practical technology tutorials and explain software concepts to a community of more than 185,000 followers.
06 / CONTACT
I'm always interested in thoughtful conversations about data, intelligent systems, and research collaborations.
[email protected]