I projektet detekteras och följs granplantor i spatiospektrala bilder för att därefter skapa en hyperspektral datakub för av varje gran. För att detektera granarna prövas fyra metoder: manuell detektion, detektion med segmentering, detektion med SVM och detektion med neuralt nätverk. Minnesanvändning och körningstid jämförs mellan två implementationer, där hyperspektral rekonstruktion görs med olika metoder.

I projektet detekteras och följs granplantor i spatiospektrala bilder för att därefter skapa en hyperspektral datakub för av varje gran. För att detektera granarna prövas fyra metoder: manuell detektion, detektion med segmentering, detektion med SVM och detektion med neuralt nätverk. Minnesanvändning och körningstid jämförs mellan två implementationer, där hyperspektral rekonstruktion görs med olika metoder.

L’objet de cette thèse est l’étude de la localisation et de la navigation à l’intérieur de bâtiments à l’aide des signaux disponibles dans les systèmes mobiles cellulaires et, en particulier, les signaux GSM.Le système GPS est aujourd’hui couramment utilisé en extérieur pour déterminer la position d’un objet, mais les signaux GPS ne sont pas adaptés à la localisation en intérieurIci, la localisation en intérieur est obtenue à partir de la technique des «empreintes» de puissance des signaux reçus sur les canaux utilisés par les réseaux GSM. Elle est réalisée à l’échelle de la pièce. Une classification est effectuée à partir de machines à vecteurs supports et les descripteurs utilisés sont les puissances de toutes les porteuses GSM. D’autres capteurs physiques disponibles dans les téléphones portables fournissent des informations utiles pour déterminer la position ou le déplacement de l’utilisateur. Celles-ci, ainsi que la cartographie de l’environnement, sont associées aux résultats obtenus à partir des «empreintes» GSM au sein de filtres particulaires afin d’obtenir une localisation plus précise, et sous forme de coordonnées continues.Les résultats obtenus montrent que l’utilisation des seules empreintes GSM permet de déterminer la pièce correcte dans 94% des cas sur une durée courte et que les performances restent stables pendant plusieurs mois, de l’ordre de 80%, si les données d’apprentissage sont enregistrées sur quelques jours. L’association de la cartographie du lieu et des informations issues des autres capteurs aux données de classification permettent d’obtenir les coordonnées de la trajectoire du système mobile avec une bonne précision et une bonne fiabilité. / GPS has long been used for accurate and reliable outdoor localization, but it cannot operate in indoor environments, which suggests developing indoor localization methods that can provide seamless and ubiquitous services for mobile users.In this thesis, indoor localization is realized making use of received signal strength fingerprinting technique based on the existing GSM networks. A room is defined as the minimum location unit, and support vector machine are used as a mean to discriminate the rooms by classifying received signal strengths from very large number of GSM carriers. At the same time, multiple sensors, such as accelerometer and gyroscope, are widely available for modern mobile devices, which provide additional information that helps location determination. The hybrid approach that combines the GSM fingerprinting results with mobile sensor and building layout information using a particle filter provides a more accurate and fine-grained localization result.The results of experiments under realistic conditions demonstrate that correct room number can be obtained 94% of the time provided the derived model is used before significant received signal strength drift sets in. Furthermore, if the training data is sampled over a few days, the performance can remain stable exceeding 80% over a period of months, and can be further improved with various post-processing techniques. Moreover, including the mobile sensors allows the system to localize the mobile trajectory coordinates with high accuracy and reliability.

No description available.

Increased strain on incumbent wastewater distribution networks originating from population increases as well as climate change calls for enhanced resource utilization. Accurately being able to predict network interconnectivity is vital within the wastewater industry to enable operational management strategies that optimizes the performance of the wastewater system. In this thesis, an evaluation of the network interconnectivity prediction performance of two machine learning models, the multilayer perceptron (MLP) and the support vector machine (SVM), utilizing supervisory control and dataacquisition (SCADA) system data for a wastewater system is presented. Results of the thesis imply that the MLP achieves the best predictions of the network interconnectivity. The thesis concludes that the MLP is the superior model and that the highest achievable network interconnectivity accuracy is 56% which is attained by the MLP model. / Den ökade påfrestningen på nuvarande avloppsnät till följd av befolkningstillväxt och klimatförändringar medför att det finns behov för optimerad resursförbrukning. Att korrekt kunna predicera ett avloppsnät är önskvärt då det möjliggör för effektivitetshöjande operativ förvaltning av avloppssystemet. I denna avhandling evalueras hur väl två maskininlärningsmodeller kan predicera nätverketssammankoppling med data från ett system för övervakning och kontroll av data (SCADA) genererat av ett avloppsnätverk. De två modellerna som testas är en multilagersperceptron (MLP) och en stödvektormaskin (SVM). Resultaten av avhandlingen visar på att MLP modellen uppnår den bästa prediktionen av nätverketssammankoppling. Avhandlingen konkluderar att MLP modellen är den bästa modellen för att predicera nätverkets sammankoppling samt att den högsta nåbara korrektheten var 56% vilket uppnåddes av MLP modellen.

Pest inventory of a field is a way of knowing when the thresholds for pest controlis reached. It is of increasing interest to use machine learning to automate thisprocess, however, many challenges arise with detection of small insects both intraps and on plants.This thesis investigates the prospects of developing an automatic warning system for notifying a user of when certain pests are detected in a trap. For this, sliding window with histogram of oriented gradients based support vector machinewere implemented. Trap detection with neural network models and a check sizefunction were tested for narrowing the detections down to pests of a certain size.The results indicates that with further refinement and more training images thisapproach might hold potential for fungus gnat and rape beetles.Further, this thesis also investigates detection performance of Mask R-CNNand YOLOv5 on different insects in fields for the purpose of automating thedata gathering process. The models showed promise for detection of rape beetles. YOLOv5 also showed promise as a multi-class detector of different insects,where sizes ranged from small rape beetles to larger bumblebees.

Elevated pulmonary artery pressure (PAP) is a significant healthcare risk. Continuous monitoring for patients with elevated PAP is crucial for effective treatment, yet the most accurate method is invasive and expensive, and cannot be performed repeatedly. Noninvasive methods exist but are inaccurate, expensive, and cannot be used for continuous monitoring. We present a machine learning model based on heart sounds that estimates pulmonary artery pressure with enough accuracy to exclude an invasive diagnostic operation, allowing for consistent monitoring of heart condition in suspect patients without the cost and risk of invasive monitoring. We conduct a greedy search through 38 possible features using a 109-patient cross-validation to find the most predictive features. Our best general model has a standard estimate of error (SEE) of 8.28 mmHg, which outperforms the previous best performance in the literature on a general set of unseen patient data.

Prenumerationstjänster blir alltmer populära i dagens samhälle. En av nycklarna för att lyckas med en prenumerationsbaserad affärsmodell är att minimera kundbortfall (eng. churn), dvs. kunder som avslutar sin prenumeration inom en viss tidsperiod. I och med den ökande digitaliseringen, är det nu enklare att samla in data än någonsin tidigare. Samtidigt växer maskininlärning snabbt och blir alltmer lättillgängligt, vilket möjliggör nya infallsvinklar på problemlösning. Denna rapport kommer testa och utvärdera ett försök att förutsäga kundbortfall med hjälp av maskininlärning, baserat på kunddata från ett företag med en prenumerationsbaserad affärsmodell där prenumeranten får besöka live-event till en fast månadskostnad. De maskininlärningsmodeller som användes i testerna var Random Forests, Support Vector Machines, Logistic Regression, och Neural Networks som alla tränades med användardata från företaget. Modellerna gav ett slutligt träffsäkerhetsresultat i spannet mellan 73,7 % och 76,7 %. Därutöver tenderade modellerna att ge ett högre resultat för precision och täckning gällande att klassificera kunder som sagt upp sin prenumeration än för de som fortfarande var aktiva. Dessutom kunde det konstateras att de kundegenskaper som hade störst inverkan på klassifikationen var ”Använda Biljetter” och ”Längd på Prenumeration”. Slutligen kommer det i denna rapport diskuteras hur informationen angående vilka kunder som sannolikt kommer avsluta sin prenumeration kan användas ur ett mer affärsmässigt perspektiv. / In today’s world subscription-based online services are becoming increasingly popular. One of the keys to success in a subscription-based business model is to minimize churn, i.e. customer canceling their subscriptions. Due to the digitalization of the world, data is easier to collect than ever before. At the same time machine learning is growing and is made more available. That opens up new possibilities to solve different problems with the use of machine learning. This paper will test and evaluate a machine learning approach to churn prediction, based on the user data from a company with an online subscription service letting the user attend live shows to a fixed price. To perform the tests different machine learning models were used, both individually and combined. The models were Random Forests, Support Vector Machines, Logistic Regression and Neural Networks. In order to train them a data set containing either active or churned users was provided. Eventually the models returned accuracy results ranging from 73.7 % to 76.7 % when classifying churners based on their activity data. Furthermore, the models turned out to have higher scores for precision and recall for classifying the churners than the non-churners. In addition, the features that had the most impact on the model regarding the classification were Tickets Used and Length of Subscription. Moreover, this paper will discuss how churn prediction can be used from a business perspective.

Quick and accurate detection of flame inside a gas turbine is very crucial to mitigaterisks in power generation. Failure of flame detection increases downtime and maintenancecosts and on rare occasions it may cause explosions due to buildup of incombustible fuel inside the combustion chamber.The aim of this thesis is to investigate the applicability ofmachine learning methods to detect the presence of flame within a gas turbine. Traditionally,this is done using an optical flame detection which converts the infrared radiation toa differential reading, which is further converted as a digital signal to the control systemand gives the flame status (1 for flame ON and 0 for flame OFF). The primary purpose ofthis alternative flame detection method is to reduce the instrument cost per gas turbine. Amachine learning model is trained with the data collected over several runs of the turbineengine and would estimate if there is an occurrence of the flame, to decide if the machineshould be ON or OFF. To reduce the instrumentation cost, the presented flame predictionmethod based on deep learning methods is employed, which takes standard data such as dynamic pressure and temperature values as input. These variables are observed to have a high correlation with the flame status. The pressure is measured using a piezocryst sensorand the temperature is measured using a thermocouple. A Study is performed by trainingon several machine learning models and coming up with which model among them have worked the best on this data.The Logistic is used as a baseline and is compared with othermodels such as KNN,SVM,Naïve Bayes,RandomForest and XGBoost is trained with thedata collected over several runs of the turbine and tested on to predict flame status insidethe gas turbine.It was observed that KNN and Random Forest performed exceptionallywell as compared to the baseline model. It is recorded that the minimum time for estimation of the flame status by the machine is 0.6 seconds and if the model implementedcan give a high accuracy with the same time then the proposed method can be an effective alternate flame detection method.

The development of technologies for preventing drowsiness at the wheel is a major challenge in the field of accident avoidance systems. Preventing drowsiness during driving requires a method for accurately detecting a decline in driver alertness and a method for alerting and refreshing the driver. As a detection method, the authors have developed a system that uses image processing technology to analyse images of the road lane with a video camera integrated with steering wheel angle data collection from a car simulation system. The main contribution of this study is a novel algorithm for drowsiness detection and tracking, which is based on the incorporation of information from a road vision system and vehicle performance parameters. Refinement of the algorithm is more precisely detected the level of drowsiness by the implementation of a support vector machine classification for robust and accurate drowsiness warning system. The Support Vector Machine (SVM) classification technique diminished drowsiness level by using non intrusive systems, using standard equipment sensors, aim to reduce these road accidents caused by drowsiness drivers. This detection system provides a non-contact technique for judging various levels of driver alertness and facilitates early detection of a decline in alertness during driving. The presented results are based on a selection of drowsiness database, which covers almost 60 hours of driving data collection measurements. All the parameters extracted from vehicle parameter data are collected in a driving simulator. With all the features from a real vehicle, a SVM drowsiness detection model is constructed. After several improvements, the classification results showed a very good indication of drowsiness by using those systems. / Title page is not included.