Ali Al-Ghaithi, MSc

The opioid epidemic has been an issue in the United States for many years now. Medical providers have become more hesitant on prescribing opioids, but there is still a large presence of heroin on the street. The purpose of this research is to find which opioid is the most abused. If it is found to be an opioid that was prescribed, how often is it prescribed, and in what dosage. Data from each state was analyzed to see which has the most opioid-related overdoses and deaths. Additionally, we want… Show more

The opioid epidemic has been an issue in the United States for many years now. Medical providers have become more hesitant on prescribing opioids, but there is still a large presence of heroin on the street. The purpose of this research is to find which opioid is the most abused. If it is found to be an opioid that was prescribed, how often is it prescribed, and in what dosage. Data from each state was analyzed to see which has the most opioid-related overdoses and deaths. Additionally, we want to see which demographic is most likely to abuse opioids. One opioid in particular, fentanyl, has recently garnered more attention due to its lethality. This data could be extremely useful to law enforcement, health care providers, and legislators. This study hopes to show the dangers of opioid abuse and trends over time. Show less

Each year, an estimated 1.3 million people die in car accidents around the world, and 20 to 50 million more sustain injuries. In this project, we explore factors that affect road transportation safety. We examined correlations between road casualties and four contributing factors: speed limit laws, vehicle safety standards, transportation infrastructure spending, and road congestion. Our analysis shows that, the number of vehicle safety standards adopted by a country has a significant impact on… Show more

Each year, an estimated 1.3 million people die in car accidents around the world, and 20 to 50 million more sustain injuries. In this project, we explore factors that affect road transportation safety. We examined correlations between road casualties and four contributing factors: speed limit laws, vehicle safety standards, transportation infrastructure spending, and road congestion. Our analysis shows that, the number of vehicle safety standards adopted by a country has a significant impact on the country’s road mortality rates. Annual road infrastructure investment and traffic congestion are somewhat significant and the maximum speed limit is not significant. Show less

Built EfficientNetB3 and CNN models to classify each cassava image into four disease categories. Created a report describing how models can help farmers increase productivity.

Examine and Applied feature engineering to the data by using different data mining techniques such as to create new important features. Created a LightGBM Molde in Python with optimized hyperparameters, combining disparate sources into one projection that outperform 1868 teams in the international Kaggle competition (Top 4%)

Operated with a group of 3, carried on the validation part to test and improve the proposed models, and helped with updating the trained model . Applied time series model such as Arima in R by creating a robust function to handle scaled calculations and presenting the reported results.

Used Python to extract data from the U.S. Bureau of Labor Statistics that contains the labor force population and unemployment rate of Nebraska, forecasting 3 months in advance

Collate 3 years of E-commerce sales data from 9 sources containing over 100K orders using Python and R. Build different ML and time series models and used AIC & MSE for model validation and predicted a three-month forecast of future sales. Determined the three best-selling categories and the fastest-growing category by using models like TSLM, OLS, and ARIMAX that were then used to forecast the next 3 months.

Built a Boosting, Random Forest, and Logistic model using the scikit-learn library to predict driver retention

Applied CatBoost model to predict the behavior of customers after visiting the website online shops. Prepared data using methods such as handling missing values, feature importance and selection, and resampling for unbalanced targets

Build a linear interpolation algorithm for determining the location and speed by second for ADS-B data

Used Pandas, BeautifulSoup, and Nump for scraping and preparing data. Prepared reports on shops and build a regression model that can predicate the sales based on the collect features. Used Matplotlib to create plots related to shops distribution and clusters.