Nowadays, the solution to many practical problems relies on machine learning tools. However, compiling the appropriate training data set for real-world classification problems is challenging because collecting the right amount of data for each class is often difficult or even impossible. In such cases, we can easily face the problem of imbalanced learning. There are many methods in the literature for solving the imbalanced learning problem, so it has become a serious question how to compare the performance of the imbalanced learning methods. Inadequate validation techniques can provide misleading results (e.g., due to data shift), which leads to the development of methods designed for imbalanced data sets, such as stratified cross-validation (SCV) and distribution optimally balanced SCV (DOB-SCV). Previous studies have shown that higher classification performance scores (AUC) can be achieved on imbalanced data sets using DOB-SCV instead of SCV. We investigated the effect of the oversamplers on this difference. The study was conducted on 420 data sets, involving several sampling methods and the DTree, kNN, SVM, and MLP classifiers. We point out that DOB-SCV often provides a little higher F1 and AUC values for classification combined with sampling. However, the results also prove that the selection of the sampler–classifier pair is more important for the classification performance than the choice between the DOB-SCV and the SCV techniques.

Road safety was always an area that concerned many people around the world and systems that aid the drivers have been appearing ever since cars and computers were combined to make driving safer and more efficient. There are plenty of systems that are able to warn drivers about different types of dangers: lane departure, collision possibility and various traffic signs. However, there is still room for development, because modern technologies, like the rising vision about the internet of things, allow us to create much more efficient systems. Also, the detections can be improved to perform better in various situations, such as different light conditions, road quality, etc. In this paper, we present the plans of a driver-assistance system, which will be capable of lane and painted traffic sign detection by using a car's on-board camera and is integrated in a network which connects different users to enhance the efficiency of these detections. The current version is capable of lane and painted traffic sign detection and is able to warn the driver about a possible lane departure on videos captured in daylight, with medium quality road markings. Our system uses a lane detection method that is based on the Hough transformation and contour detection for the painted traffic sign detection.