Parking vehicle is a daunting task and a common problem in many cities around the globe. The search for parking space leads to congestion, frustration and increased air pollution. Information of a vacant parking space would facilitate to reduce congestion and subsequent air pollution. Therefore, aim of the paper is to acquire vehicle occupancy in an open parking lot using deep learning. Thermal camera was used to collect the data during varying environmental conditions such as; sunny, dusk, dawn, dark and snowy conditions. Vehicle detection with deep learning was implemented where image classification and object localization were performed for multi object detection. The dataset consists of 527 images which were manually labelled as there were no pre-labelled thermal images available. Multiple deep learning networks such as Yolo, ReNet18, ResNet50 and GoogleNet with varying layers and architectures were evaluated on vehicle detection. Yolo, GoogleNet and ResNet18 are computationally efficient detectors which took less processing time while Resnet50 produced better detection results compared to other detectors. However, ResNet18 also produced minimal miss rates and is suitable for real time vehicle detection. The detected results were compared with a template of parking spaces and IoU value is used to identify vehicle occupancy information.

The aim of this paper is to investigate the feasibility of using the Dynamic Time Warping (DTW) method to measure motor states in advanced Parkinson’s disease (PD). Data were collected from 19 PD patients who experimented leg agility motor tests with motion sensors on their ankles once before and multiple times after an administration of 150% of their normal daily dose of medication. Experiments of 22 healthy controls were included. Three movement disorder specialists rated the motor states of the patients according to Treatment Response Scale (TRS) using recorded videos of the experiments. A DTW-based motor state distance score (DDS) was constructed using the acceleration and gyroscope signals collected during leg agility motor tests. Mean DDS showed similar trends to mean TRS scores across the test occasions. Mean DDS was able to differentiate between PD patients at Off and On motor states. DDS was able to classify the motor state changes with good accuracy (82%). The PD patients who showed more response to medication were selected using the TRS scale, and the most related DTW-based features to their TRS scores were investigated. There were individual DTW-based features identified for each patient. In conclusion, the DTW method can provide information about motor states of advanced PD patients which can be used in the development of methods for automatic motor scoring of PD.

Electricity price forecasting plays a crucial role in a liberalized electricity market. In terms of forecasting approaches, computational intelligence based models have been widely used with respect to electricity price forecasting and among all computation intelligence based models, artificial neural networks are most popular among researchers due to their flexibility and efficiency in handling complexity and non-linearity. However, a review of recent applications of neural networks for electricity price forecasting is not found in the literature. The motivation of this paper is to fill this research gap. In this study, existing approaches are analyzed and a summary of the strengths and weaknesses of each approach is presented. Besides, each neural network model is briefly summarized, followed by reviews of the corresponding studies of each neural network with respect to electricity forecasting from year 2010 onwards. Major contributions, datasets adopted as well as the corresponding experiment results are analyzed for each reviewed study. Apart from the review of existing studies, the advantages and disadvantages of each type of neural network model are discussed in details. Compared with neural networks based hybrid models, a single neural network model is easier to be implemented, less complex and more efficient. Scope of the review is the application of non-hybrid neural network models. It is found that most literature focuses on short term electricity price forecasting while medium and long term forecasting still remain relatively uncovered.

Workplace safety is a major concern in many countries. Among various industries, construction sector is identified as the most hazardous work place. Construction accidents not only cause human sufferings but also result in huge financial loss. To prevent reoccurrence of similar accidents in the future and make scientific risk control plans, analysis of accidents is essential. In construction industry, fatality and catastrophe investigation summary reports are available for the past accidents. In this study, text mining and natural language process (NLP) techniques are applied to analyze the construction accident reports. To be more specific, five baseline models, support vector machine (SVM), linear regression (LR), K-nearest neighbor (KNN), decision tree (DT), Naive Bayes (NB) and an ensemble model are proposed to classify the causes of the accidents. Besides, Sequential Quadratic Programming (SQP) algorithm is utilized to optimize weight of each classifier involved in the ensemble model. Experiment results show that the optimized ensemble model outperforms rest models considered in this study in terms of average weighted F1 score. The result also shows that the proposed approach is more robust to cases of low support. Moreover, an unsupervised chunking approach is proposed to extract common objects which cause the accidents based on grammar rules identified in the reports. As harmful objects are one of the major factors leading to construction accidents, identifying such objects is extremely helpful to mitigate potential risks. Certain limitations of the proposed methods are discussed and suggestions and future improvements are provided.

Patients with Parkinson disease (PD) exhibit a gait disorder called festinating gait which is caused by deficiency of dopamine in the basal ganglia. To analyze gait of patients with PD, different spatiotemporal parameters such as stride length, cadence, and walking speed should be calculated. This paper aims to present a method to extract useful information represented by the positions of certain landmarks on the human body that can be used for analysis of PD patients’ gait. This method is tested using 132 videos collected from 7 PD patients and 7 healthy controls. The positions of 4 body landmarks, namely body’s center of gravity (COG), the position of the head, and the position of the feet, was computed using a total of more than 41000 of video frames. Results of object’s movement plots show high level of accuracy in the calculation of the body landmarks.

Electricity price forecasting plays a crucial role in aliberalized electricity market. In terms of forecasting approaches,artificial neural networks are the most popular amongresearchers due to their flexibility and efficiency in handlingcomplexity and non-linearity. On the other hand, a single neuralnetwork presents certain limitations. Therefore, in recent years,hybrid models that combine multiple algorithms to balance outthe advantages of a single model have become a trend. However,a review of recent applications of hybrid neural networks basedmodels with respect to electricity price forecasting is not found inthe literature and hence, the motivation of this paper is to fill thisresearch gap. In this study, methodologies of existing forecastingapproaches are briefly summarized, followed by reviews of neuralnetwork based hybrid models concerning electricity forecastingfrom year 2015 onwards. Major contributions of each study,datasets adopted in experiments as well as the correspondingexperiment results are analyzed. Apart from the review ofexisting studies, the novelty and advantages of each type of hybridmodel are discussed in detail. Scope of the review is theapplication of hybrid neural network models. It is found that theforecast horizon of the reviewed literature is either hour ahead orday ahead. Medium and long term forecasting are notcomprehensively studied. In addition, though hybrid modelsrequire relatively large computational time, time measurementsare not reported in any of the reviewed literature.

Parking a vehicle is a daunting task during peak hours. The search for a parking space leads to congestion and increased air pollution. Information of a vacant parking space would facilitate to reduce congestion and subsequent air pollution. This paper aims to identify parking occupancy in an open parking lot which consists of free parking spaces using a thermal camera. A thermal camera is capable of detecting vehicles in any weather and light conditions based on emitted heat and it can also be installed in public places with less restrictions. However, a thermal camera is expensive compared to a colour camera. A thermal camera can detect vehicles based on the emitted heat without any illumination. Vehicles appear bright or dark based on heat emitted by the vehicles. In order to identify vehicles, pre-trained vehicle detection algorithms, Histogram of Oriented Gradient detectors, Faster Regional Convolutional Neural Network (FRCNN) and modified Faster RCNN deep learning networks were implemented in this paper. The detection rates of the detectors reduced with diminishing of heat in the vehicles. Modified Faster RCNN deep learning network produced better detection results compared to other detectors. However, the detection rates can further be improved with larger and diverse training dataset.

Electricity price forecasting plays a crucial role in liberalized electricity markets. Generally speaking, short term electricity price forecasting is essential for electricity providers to adjust the schedule of production in order to balance consumers’ demands and electricity generation. Short term forecasting results are also utilized by market players to decide the timing of purchasing or selling to gain maximized profit. Among existing forecasting approaches, neural networks are regarded as the state of art method. However, deep neural networks are not studied comprehensively in this field, thus the motivation of this study is to fill this research gap. In this paper, a novel hybrid approach is proposed for short term electricity price forecasting. To be more specific, categorical boosting (Catboost) algorithm is used for feature selection and a bidirectional long short term memory neural network (BDLSTM) serves as the main forecasting engine. To evaluate the effectiveness of the proposed approach, two datasets from the Nord Pool market are employed in the experiment. Moreover, the performance of multi-layer perception (MLP) neural network, support vector regression (SVR) and ensemble tree models are evaluated and compared with the proposed model. Results show that the proposed approach outperforms the rest models in terms of mean absolute percentage error (MAPE).