Thesis Chapters by Tedom Donald
Donald N. Tedom, 2024
Accurate prediction of solar radiation is crucial for enhancing the efficiency of renewable energ... more Accurate prediction of solar radiation is crucial for enhancing the efficiency of renewable energy systems, agriculture, and climate studies, leading to sustainable energy solutions. This project aims to find the best machine learning and deep learning models for predicting solar radiation across ten regions in Cameroon. The models used include Linear Regression (LR), Decision Tree (DT), Random Forest (RF), Gradient Boosting Machine (GB), K-Nearest Neighbours (KNN), Long Short-Term Memory (LSTM), Artificial Neural Network (ANN), Feedforward Neural Network (FNN), Recurrent Neural Network (RNN), and Convolutional Neural Network (CNN). Using data spanning 20 years from NASA's MERRA-2 (2002-2021) and assessing metrics such as R² (0.720 to 0.957), RMSE (0.246 to 1.498 KJ/m²), rRMSE, MBE, MABE (0.131 to 1.104 KJ/m²), t-statistic, MAE, and MSE, the study found significant variability in model performance. Based on these metrics, KNN and RF were the best-performing machine learning models, while CNN was the best among deep learning models, providing reliable predictions with lower errors. For making solar radiation predictions in Cameroon, KNN, RF, and CNN are the recommended models to use.
Keywords: Solar Radiation Forecasting, Machine Learning, Cameroon, Renewable Energy, Model Comparison, Prediction Accuracy.
Tedom Noutchogouin Donald , 2023
In recent years, ecommerce applications in Cameroon have grown significantly, resulting in a wide... more In recent years, ecommerce applications in Cameroon have grown significantly, resulting in a wide array of products. However, improving user experience remains a major challenge for existing platforms, particularly with their categorical-based recommendation systems. These systems recommend products based on user purchases within a specific category, which has proven to be inefficient. With the rise of machine learning, alternative solutions offering improved performance have emerged. This report presents the development and evaluation of an ecommerce application with a machine learning-based recommendation system. The project aims to enhance personalized shopping experiences, increase company revenue, and improve customer satisfaction through accurate and relevant product recommendations. The recommendation system employs collaborative filtering, neural networks, content-based, and knowledge-based methods, trained on a dataset of 10,000 records from GroupLens capturing product ratings by hundreds of users in Cameroon. The evaluation indicates promising accuracy (60%), mean reciprocal rank (MRR) of 0.26, mean absolute error (MAE) of 0.214, root mean squared error (RMSE) of 0.370, precision of 0.8, and recall of 0.8. Recently, the ecommerce app was hosted at https://hooyia-market.onrender.com, attracting 13 new user registrations within a week. Where 45 products were added to the cart, with 15 directly influenced by the recommendation system, representing 33.3% of total cart additions. Additionally, the recommendation-driven purchases may lead to a 33.3% increase in sales. The findings demonstrate the hybrid recommendation system's effectiveness in providing accurate and relevant recommendations, significantly enhancing shopping experiences and user engagement in Cameroon. Overall, this project contributes to the field of ecommerce and machine learning, showcasing a successful implementation and evaluation of a hybrid recommendation system tailored for Cameroon. The results serve as a valuable reference for ecommerce companies seeking similar approaches to enhance customer experiences, increase revenue, and foster business growth.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.
Donald Tedom, 2023
In recent years, ecommerce applications in Cameroon have grown significantly, resulting in a wide... more In recent years, ecommerce applications in Cameroon have grown significantly, resulting in a wide array of products. However, improving user experience remains a major challenge for existing platforms, particularly with their categorical-based recommendation systems. These systems recommend products based on user purchases within a specific category, which has proven to be inefficient. With the rise of machine learning, alternative solutions offering improved performance have emerged. This report presents the development and evaluation of an ecommerce application with a machine learning-based recommendation system. The project aims to enhance personalized shopping experiences, increase company revenue, and improve customer satisfaction through accurate and relevant product recommendations. The recommendation system employs collaborative filtering, neural networks, content-based, and knowledge-based methods, trained on a dataset of 10,000 records from GroupLens capturing product ratings by hundreds of users in Cameroon. The evaluation indicates promising accuracy (60%), mean reciprocal rank (MRR) of 0.26, mean absolute error (MAE) of 0.214, root mean squared error (RMSE) of 0.370, precision of 0.8, and recall of 0.8. Recently, the ecommerce app was hosted at https://hooyia-market.onrender.com, attracting 13 new user registrations within a week. Where 45 products were added to the cart, with 15 directly influenced by the recommendation system, representing 33.3% of total cart additions. Additionally, the recommendation-driven purchases may lead to a 33.3% increase in sales. The findings demonstrate the hybrid recommendation system's effectiveness in providing accurate and relevant recommendations, significantly enhancing shopping experiences and user engagement in Cameroon. Overall, this project contributes to the field of ecommerce and machine learning, showcasing a successful implementation and evaluation of a hybrid recommendation system tailored for Cameroon. The results serve as a valuable reference for ecommerce companies seeking similar approaches to enhance customer experiences, increase revenue, and foster business growth.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.
Uploads
Thesis Chapters by Tedom Donald
Keywords: Solar Radiation Forecasting, Machine Learning, Cameroon, Renewable Energy, Model Comparison, Prediction Accuracy.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.
Keywords: Solar Radiation Forecasting, Machine Learning, Cameroon, Renewable Energy, Model Comparison, Prediction Accuracy.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.
Keywords: Ecommerce, Recommendation System, Hybrid Models, Cameroon, Sales Impact, Customer Satisfaction.