KRISTINA DIBELLA-SILVA

Committed job seeker with a history of meeting and exceeding company needs with consistent and organized practices. In-depth understanding of customer relations and project management. Analytical and critical thinker with a remarkable work ethic. Detail-oriented and able to learn new concepts quickly. Skilled in working under pressure and adapting to new situations and challenges to best enhance the organization.

Adult Income Prediction 

Building a Classification prediction model using: KNN,  Logistic Regression, Decision Tree Classifier & Random Forest Classifier

• Extracted, processed and analyzed the dataset that contains various income variables, such as the person's education, employment field, gender and ethnicity.
• Feature engineered parts of the data to minimize noise in our models using python code.
• Analyzed which categories of this dataset have the most influence in our predictions using pandas and multiple data visualizations.
• Evaluated the effects of using PCA(Principal Component Analysis) in our model pipeline to further filter the most efficient amount of data.
• Tuned a Decision Tree model with Undersampling to evaluate the results as this dataset was significantly imbalanced with the majority making less than 50K.
• Achieved 85.7% accuracy in predicting which income group an individual will be.
• Presented powerpoint to stakeholders on final analysis of this dataset.

Sales Predictions Model

Building a sales prediction model using: Linear Regression (Lasso), Decision Tree Regressor

• Extracted, processed and analyzed the dataset that contains historical sales information from 10 Big Mart stores
• Developed a predictive model that can help retailers determine the number of sales they are going to make in the future to meet customer’s demands and expectations
• This final model provided a R2 score of 61% where our target variance was accounted for in its predictions.