SOLOMON AMOO

am a Ph.D student in data science, equipped with relevant skillset in data analytics and Machine Learning.

• Aws Certified Solutions Architect-Associate
• Aws Certified Machine Learning-Specialty

• Monte Carlo method for option pricing - Implemented the Monte Carlo method for option pricing, improving accuracy by employing a control variate approach to mitigate variance and reduce standard deviation under stochastic volatility models (Hull and White, Stein-Stein, Heston). Incorporated auxiliary processes to transform stochastic volatility into a deterministic form, resulting in enhanced simulation performance and reduced uncertainty in option price estimates.
• Sentiment analysis on stock data - Conducted sentiment analysis on stock data using artificial intelligence techniques, specifically implementing a deep neural network model. Analyzed and interpreted stock sentiment patterns to extract valuable insights.
• Bank market segmentation - Conducted bank market segmentation through unsupervised machine learning techniques, specifically K-Means clustering. Utilized Principal Component Analysis (PCA) and Autoencoders for dimensionality reduction, enhancing feature representation in the data.
• Credit card fraud forecasting - Utilized machine learning techniques to forecast credit card fraud. Due to the skewed data distribution, the SMOTE, Oversampling and Undersampling technique is employed to balance the dataset. A combination of voting classifiers is implemented, while hyperparameters are fine-tuned using both GridSearchCV and RandomSearchCV. To accurately assess model performance given the class imbalance, the Area Under the Precision-Recall Curve (AUPRC) metric is adopted. The models used includes Convolutional Neural network Recurrent Neural Network, Long Short-Term Memory, K-Nearest Neighbors (KNN), Logistic Regression, Random Forest, XGBoost, Support Vector Classifier (SVC), and Decision Tree Classifier.
• Binary classification of heart disease - Developed a machine learning model for binary classification of heart disease using medical attributes. Various models including Logistic Regression, K-Nearest Neighbors, and Random Forest were trained and compared based on their predictive performance. The goal was to determine whether a person has heart disease using different features.
• Bulldozer sale price prediction - Worked on a machine learning project to predict bulldozer sale prices. Employed the Random Forest Regressor model for prediction and evaluated model performance using the Root Mean Squared Error (RMSE).