Parkinson's disease (PD) is a neurological condition that affects the brain of the human body and causes difficultywalking, shaking, stiffness, and loss of balance and coordination. Most of the patients suffering from PD face challenges in speaking during the initial stages. In this study, illness has been classified by applying speech features. The standard speech components employed in Parkinson's Disease are Shimmer, Jitter, Harmonic parameters, Frequency parameters, Detrended Fluctuation Analysis (DFA), Recurrence Period Density Entropy (RPDE), and Pitch Period Entropy (PPE) (PD). These features are the baseline features chosen for this study. CNN and XGBoost have been selected to classify the model andrecognize Parkinson's Disease in the early stages. From the model feature, the selection was excluded to improve the model.
The pH value of soil is a critical parameter which governs the fertility of soil. In Uttarakhand, due to very encouraging climatic conditions like temperature, moisture and humidity, the soil has an appropriate pH value, and hence it is very fertile, due to which, the agriculture has become a significant source of income for Uttarakhand people. This research article is based upon the secondary data pertaining to soil of Kumaun region of Uttarakhand. This available data has been used to develop and train RBF NN (radial basis function neural network) and FIS (subtractive clustering based fuzzy inference system), on MATLAB, for predicting the pH of soil out of the given minerals concentration e.g. P, K, Fe, Mn, Cu. Eventually, the prediction performance of both the designed models are compared in terms of mean square prediction error.
This research focuses its objectives on the use of machine learning techniques in the detection of images of a CT scan to diagnose Pancreatic cancer. The work used an extensive database containing diverse CT images, more specifically, concerning pancreatic cancer patients. Two primary methodologies were implemented in this study. The first was the Support Vector Machine (SVM) classifier, one of the most used classifier methods in classification problems. The second component is the Convolution Neural Network (CNN), which has ResNet 50 at its base. The architecture of this model is commendably profound, and it can perceive the details of the images. The SVM classifier marked a precision of 83.07% for detecting cancer; such a level of accuracy can give clues, though it is still possible to define additional possibilities for improving the classification of complicated cases. On the other hand, the CNN model based on the ResNet50 learned a more accurate model than the training accuracy of 90. 01%. Such a performance increase is explained quite well by deep learning methodologies’ capabilities, especially in medical image analysis. These studies draw attention to the possibility of utilizing more sophisticated CNN models to promote the accuracy and reliability of pancreatic cancer diagnosis. The increased accuracy also has ramifications for the diagnostic phase of pancreatic cancer, but also it emphasizes the necessity of using complex algorithms for early identification of this disease.
This research evaluates four Deep Reinforcement Learning (DRL) algorithms-TD3, SAC, Rainbow DQN, and TRPO-for their effectiveness in adaptive traffic management, specifically in optimizing parking space management in smart cities. The algorithms were assessed based on key performance metrics: TD3 achieved a reward score of 85.6 with Actor Loss of 0.023, Critic Loss of 0.041, and 72 hours of training time. SAC outperformed with a reward score of 88.2, Actor Loss of 0.018, Critic Loss of 0.035, and 68 hours of training. Rainbow DQN achieved a competitive reward score of 87.4, Q-value of 0.075, Loss of 0.027, and 80 hours of training. TRPO attained a reward score of 86.8, Policy Loss of 0.032, Value Function Loss of 0.021, and 75 hours of training. SAC has shown best performance across the board, achieving a good exploration and exploitation trade-off. While Rainbow DQN can be seen excelling on tasks with discrete action spaces, the TD3 and TRPO clearly outperformed the others from continuous control task performance and a robust stable policy update of following iterations perspective respectively. The results show that SAC had the best overall performance with demonstrating a good trade-off between exploration and exploitation. While Rainbow DQN was robust to dilution in discrete action spaces, TD3 and TRPO performed well on continuous control tasks and stable policy updates respectively. These results give further evidence to the potential of DRL for real-world applications, such as dynamically assigning parking slots and directing traffic in cities towards efficient city-wide mobility. Future research may examine the further integration of real-time IoT data and increased scalability for citywide implementations to deliver smarter, more adaptive, traffic management solutions that account for the complexities of today's urban areas.
Image classification, also known as object detection, is an advanced form of computer vision in which a network can predict objects in an image and point them out as bounding boxes. Classifying and recognizing objects in an image is just one aspect of object detection because it also involves localizing these objects and attracting bounding boxes around them. Since object detection involves image and video analysis, it has been of critical interest to researchers over the past few years. The previous object recognition method relies on manually created features and traditional and trainable algorithms. By using the YOLO (You Only Look Once) process, this research aims to perform real-time object detection. Regarding speed and performance, the YOLO algorithm is much more efficient than the existing model. Unlike existing algorithms that do not scan the entire image in single forward propagation, YOLO uses a neural network to predict binding boxes for the whole image. In comparison with existing algorithms, YOLO performs faster.