Vol. 2 No. 4 (2023)

Nowadays, it is crucial to transfer official documents such as invoices, dispatch notes, and receipts into digital environments and establish correct semantic relationships. However, understanding and processing these documents is a difficult process that requires significant time and effort. In recent years, the use of deep learning, image preprocessing, text detection, and optical character recognition (OCR) technologies have made this process easier. However, for text recognition and processing techniques to produce accurate results, documents must be clean and readable. Additionally, difficulties arising from time-consuming, tiring, error-prone, and cost-incurring human-powered digitalization processes must be reduced. The aim of this study is to digitize and archive scanned invoices and similar official documents using current artificial intelligence technologies, thereby enabling the most effective use of components such as time, cost, and human resources. The dataset used in the study includes 10,000 ".jpg" image files and 10,000 ".xml" data files. The model trained with the ResNet-50 architecture can detect text with accuracy rates of up to 97% on randomly selected images from the dataset. In an environment where a person can process an average of 2,112 documents per month, it is predicted that the trained artificial intelligence model can process 108,000 documents per month. With this developed method, businesses can quickly digitize and archive official documents such as invoices, dispatch notes, and receipts. Future studies propose the development of new methods that can produce better results using larger and more diverse datasets.

In this study, the potential of machine learning methods for analyzing future needs in the logistics sector was investigated. The research is conducted using the MATLAB platform. Numeric pallet demand data obtained from a logistics company are employed to train MLP, LSTM, and CNN models. Data security and confidentiality take priority during the data collection process. This dataset, comprising a total of 3,062 daily records, serves as the primary data source for the study. In the data preprocessing phase, missing or erroneous data is rectified, and outliers are detected and corrected. The models are tested to predict pallet quantities over periods of 25 days and 4 weeks. The results are evaluated by comparing the model predictions with actual data. Model performances are assessed using metrics such as MSE, RMSE, NRMSE, MAE, ESD, and RC. The outcomes of the last 25 days demonstrate that the LSTM model exhibits the lowest MSE (6,410.5571) and RMSE (80.0660) values. For the MLP model, the MSE value is calculated as 20,536.5564, and the RMSE value is 143.3058. Performance evaluations for the CNN model yield an MSE of 8,492.4297 and an RMSE of 92.1544. Furthermore, it is observed that the MLP model provides the best results for the 4-week forecasts. The results of this study indicate the success of the models used for predicting pallet transportation quantities in the logistics sector. In addition to this study, a contribution is made toward enabling logistics companies to make more informed and strategic decisions.

Hearts are crucial for maintaining a healthy lifestyle and are ranked high among the organs that need special care. Globally, heart disease is one of the leading causes of death, posing a considerable public health challenge in low-income countries in particular. Early diagnosis and the identification of risk factors are critical when dealing with these diseases, as early symptoms are often not evident. Heart disease can be caused by several factors, including smoking, poor diet, stress, a lack of physical activity, and excessive alcohol consumption. During the diagnosis process, doctors may encounter various challenges, including vague symptoms, misleading test results, and other medical complications. It is currently possible to diagnose heart disease more accurately and effectively using machine learning algorithms. The present study examines seven different machine learning algorithms on a dataset consisting of 4,238 records and 16 different patient characteristics. Among the classification models, Naive Bayes, Decision Trees, Random Forests, Support Vector Machines (SVM), Artificial Neural Networks (ANNs), K Nearest Neighbors, and Logistic Regressions yielded 78.9%, 79.9%, 83.9%, 70.9%, 83.7%, 83.4%, and 85.5%, respectively.