Cervical cancer is a major health concern for women around the world, especially in countries where there is limited access to regular Pap smear screenings. A Pap smear test can help detect cervical cancer early. However, analyzing the images manually is slow and labor intensive. Convolutional Neural Networks (CNNs) show promise in classifying Pap smear test images quicker and more accurately. CNNs are a unique type of artificial intelligence that excel at recognizing patterns, shapes, and features in pictures. However, this model needs a significant amount of data to work well. One major challenge in training these models is this data imbalance; there are many more images of healthy cells than unhealthy. This imbalance creates a problem because the model can become biased towards predicting healthy cells. To improve the fairness and accuracy of CNN-based models, we explored different ways to manage the data imbalance using published datasets of healthy and unhealthy human cells. Three methods were tested: 1) creating more images of cancer cells through image augmentation to increase the number of class images; 2) designing random oversampling and undersampling to balance the dataset to add more samples to minority classes or reducing samples from majority classes; and, 3) utilizing Class Weight to increase the loss penalty for underrepresented classes.

Artificial Intelligence is revolutionizing the financial field, such as how we manage money, measure risk, and make decisions. For many years, financial institutions always relied on traditional models and human intuition to make the best decisions. While these methods were effective, they were not as convenient as AI in recent years since you can make financial decisions by simply typing a sentence and clicking a button. This research project will use the quantitative and qualitative analysis methods, looking at different programs being created and how they can be used for making profit trading in real time. These programs also have the advantage of optimizing your portfolio, measuring risk, detecting fraud and so many more applications. The anticipated findings for this research project will include real life examples of AI in trading, as well as firms and individuals using it to enhance efficiency and highlight the bigger platforms of the study. Another aim for this study is to show the challenges regarding transparency and examples of data bias and ethical concerns. The purpose of this research is to understand both the potential and the limitations of AI, specifically in finance while emphasizing the importance of responsible use to make sure there is sustainability in financial markets with this new technology.