This thesis presents detailed observation of different Convolutional Neural Network (CNN) architecture which had assisted Computer Vision researchers to achieve state-of-the-art performance on classification, detection, segmentation and much more to name image analysis challenges. Due to the advent of deep learning, CNN had been used in almost all the computer vision applications and that is why there is utter need to understand the miniature details of these feature extractors and find out their pros and cons of each feature extractor meticulously. In order to perform our experimentation, we decided to explore an object detection task using a particular model architecture which maintains a sweet spot between computational cost and accuracy. The model architecture which we had used is LSTM-Decoder. The model had been experimented with different CNN feature extractor and found their pros and cons in variant scenarios. The results which we had obtained on different datasets elucidates that CNN plays a major role in obtaining higher accuracy and we had also achieved a comparable state-of-the-art accuracy on Pedestrian Detection Dataset. In extension to object detection, we also implemented two different model architectures which find shoulder keypoints. So, One of our idea can be explicated as follows: using the detected annotation from object detection, a small cropped image is generated which would be feed into a small cascade network which was trained for detection of shoulder keypoints. The second strategy is to use the same object detection model and fine tune their weights to predict shoulder keypoints. Currently, we had generated our results for shoulder keypoint detection. However, this idea could be extended to full-body pose Estimation by modifying the cascaded network for pose estimation purpose and this had become an important topic of discussion for the future work of this thesis.

Long short-term memory (LSTM) neural networks have been proven to be effective for time series prediction, even in some instances where the data is non-stationary. This lead us to examine their predictive ability of stock market returns, as the development of stock prices and returns tend to be a non-stationary time series. We used daily stock trading data to let an LSTM train models at predicting daily returns for 60 stocks from the OMX30 and Nasdaq-100 indices. Subsequently, we measured their accuracy, precision, and recall. The mean accuracy was 49.75 percent, meaning that the observed accuracy was close to the accuracy one would observe by randomly selecting a prediction for each day and lower than the accuracy achieved by blindly predicting all days to be positive. Finally, we concluded that further improvements need to be made for models trained by LSTMs to have any notable predictive ability in the area of stock returns.

No description available.

Accurate forecasts are vital in multiple areas of economic, scientific, commercial, and industrial activity. There are few previous studies on using forecasting methods for predicting serious events. This thesis set out to investigate two things, firstly whether machine learning models could be applied to the objective of forecasting serious events. Secondly, if the models could be made interpretable. Given these objectives, the approach was to formulate two forecasting tasks for the models and then use the Python framework SHAP to make them interpretable. The first task was to predict if a serious event will happen in the coming eight hours. The second task was to forecast how many serious events that will happen in the coming six hours. GBDT and LSTM models were implemented, evaluated, and compared on both tasks. Given the problem complexity of forecasting, the results match those of previous related research. On the classification task, the best performing model achieved an accuracy of 71.6%, and on the regression task, it missed by less than 1 on average. / Exakta prognoser är viktiga inom flera områden av ekonomisk, vetenskaplig, kommersiell och industriell verksamhet. Det finns få tidigare studier där man använt prognosmetoder för att förutsäga allvarliga händelser. Denna avhandling syftar till att undersöka två saker, för det första om maskininlärningsmodeller kan användas för att förutse allvarliga händelser. För det andra, om modellerna kunde göras tolkbara. Med tanke på dessa mål var metoden att formulera två prognosuppgifter för modellerna och sedan använda Python-ramverket SHAP för att göra dem tolkbara. Den första uppgiften var att förutsäga om en allvarlig händelse kommer att ske under de kommande åtta timmarna. Den andra uppgiften var att förutse hur många allvarliga händelser som kommer att hända under de kommande sex timmarna. GBDT- och LSTM-modeller implementerades, utvärderades och jämfördes för båda uppgifterna. Med tanke på problemkomplexiteten i att förutspå framtiden matchar resultaten de från tidigare relaterad forskning. På klassificeringsuppgiften uppnådde den bäst presterande modellen en träffsäkerhet på 71,6%, och på regressionsuppgiften missade den i genomsnitt med mindre än 1 i antal förutspådda allvarliga händelser.

The space of hand gesture recognition using radar and sonar is dominated mostly by radar applications. In addition, the machine learning algorithms used by these systems are typically based on convolutional neural networks with some applications exploring the use of long short term memory networks. The goal of this study was to build and design a Sonar system that can classify hand gestures using a machine learning approach. Secondly, the study aims to compare convolutional neural networks to long short term memory networks as a means to classify hand gestures using sonar. A Doppler Sonar system was designed and built to be able to sense hand gestures. The Sonar system is a multi-static system containing one transmitter and three receivers. The sonar system can measure the Doppler frequency shifts caused by dynamic hand gestures. Since the system uses three receivers, three different Doppler frequency channels are measured. Three additional differential frequency channels are formed by computing the differences between the frequency of each of the receivers. These six channels are used as inputs to the deep learning models. Two different deep learning algorithms were used to classify the hand gestures; a Doppler biLSTM network [1] and a CNN [2]. Six basic hand gestures, two in each x- y- and z-axis, and two rotational hand gestures are recorded using both left and right hand at different distances. The gestures were also recorded using both left and right hands. Ten-Fold cross-validation is used to evaluate the networks' performance and classification accuracy. The LSTM was able to classify the six basic gestures with an accuracy of at least 96% but with the addition of the two rotational gestures, the accuracy drops to 47%. This result is acceptable since the basic gestures are more commonly used gestures than rotational gestures. The CNN was able to classify all the gestures with an accuracy of at least 98%. Additionally, The LSTM network is also able to classify separate left and right-hand gestures with an accuracy of 80% and The CNN with an accuracy of 83%. The study shows that CNN is the most widely used algorithm for hand gesture recognition as it can consistently classify gestures with various degrees of complexity. The study also shows that the LSTM network can also classify hand gestures with a high degree of accuracy. More experimentation, however, needs to be done in order to increase the complexity of recognisable gestures.

An idea of storing contents, such as media files, music files, movie clips, etc. is simple yet challenging in terms of required effort to make it count. Some of the benefits of pre-storing the contents are reduced delay of accessing/downloading a content, reduced load to the centralized servers and of course, a higher data rate. However, several challenges need to be addressed to achieve these benefits. Among many, some of the fundamentals are limited storage capacity, storing the right content and minimizing the costs. This thesis aims to address these challenges. First, a framework for predicting the proper contents that need to be stored to the limited storage capacity is presented. Then, the cost is minimized considering several real-world scenarios. While doing that, all possible collaborations among the local nodes are performed to ensure high performance. Therefore, the goal of this thesis is to come up with a solution to the content storing problems so that the network cost is minimized.

No description available.

No description available.

Prediction of future stock markets has long been, and will continue to be a relevant topic. However, predicting markets is one of the most challenging areas to work with due to the unpredictability of the market. The extent to which markets can be predicted is a debated subject that has not yet been answered. A common approach is to use machine learning in combination with historical data to predict future stock prices. In this report, a classical machine learning method, LSTM, will be applied to nickel product prices to predict future product prices. The data used is provided by the company Harald Pihl, which has been trading various metals since the early 1900s. As a comparative material, the method is also applied to data on the nickel futures market. The results conclude that a larger number of data points are required for the prediction of nickel products to generate a credible result. In addition to this, there is a significant variation in the quality of the results depending on the dataset being used. The difference in results is due, among other things, to the number of data points, fluctuations in the dataset, and the regularity of the dataset.

This thesis uses a multi-tasking long short-term memory (LSTM) model to investigate the correlation between containment strategies, climate change, and the number of COVID-19 transmissions and deaths. The study focuses on examining the accuracy of different factors in predicting the number of daily confirmed cases and deaths cases to further explore the correlation between different factors and cases. The initial assessment results suggest that containment strategies, specifically vaccination policies, have a more significant impact on the accuracy of predicting daily confirmed cases and deaths from COVID-19 compared to climate factors such as the daily average surface 2-meter temperature. Additionally, the study reveals that there are unpredictable effects on predictive accuracy resulting from the interactions among certain impact factors. However, the lack of interpretability of deep learning models poses a significant challenge for real-world applications. This study provides valuable insights into understanding the correlation between the number of daily confirmed cases, daily deaths, containment strategies, and climate change, and highlights areas for further research. It is important to note that while the study reveals a correlation, it does not imply causation, and further research is needed to understand the trends of the pandemic.