京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第19109号 / 情博第555号 / 新制||情||98(附属図書館) / 32060 / 京都大学大学院情報学研究科知能情報学専攻 / (主査)教授 阿久津 達也, 教授 山本 章博, 教授 鹿島 久嗣 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20054号 / 医博第4162号 / 新制||医||1018(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 湊谷 謙司, 教授 富樫 かおり, 教授 木村 剛 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

A comparison of performance between tradition support vector machine (SVM), single kernel, multiple kernel learning (MKL), and modern deep learning (DL) classifiers are observed in this thesis. The goal is to implement different machine-learning classification system for object detection of three dimensional (3D) Light Detection and Ranging (LiDAR) data. The linear SVM, non linear single kernel, and MKL requires hand crafted features for training and testing their algorithm. The DL approach learns the features itself and trains the algorithm. At the end of these studies, an assessment of all the different classification methods are shown.

Background. Due to an increasing number of credit card defaulters, companies arenow taking greater precautions when approving credit applications. When a customermeets certain requirements, credit card firms typically use their experience todecide whether to grant them a credit card. Additionally, a few machine learningmethods have been applied to support the final decision. Objectives. The aim of this thesis is to compare the accuracy of logistic regressionclassifier, random forest classifier, and support vector classifier with the ensemblebagging classifier for predicting credit card approval. Methods. This thesis follows a method called general experimentation to determinethe most accurate classification technique for predicting credit card approval. Thedataset is taken from Kaggle, which contains information about credit card applications.The selected algorithms are trained with training data and validate themusing validation data then evaluate their performance on the testing data by usingmetrics such as accuracy, precision, recall, F1 score, and ROC curve. Now ensemblelearning bagging technique is applied to combine the predictions of these multiplemodels using majority voting to create an ensemble model. Finally, the performanceof the ensemble model was evaluated on the testing data and compared its accuracyto that of the individual models to identify the most accurate classification techniquefor predicting credit card approval. Results. Among the four selected machine learning algorithms, the random forestclassifier performed better with an accuracy of 88.41% on the testing dataset.The second-best algorithm is the ensemble bagging classifier, with an accuracy of84.78%. Hence, the random forest classifier is the most accurate algorithm for predictingcredit card approval. Conclusions. After evaluating various classifiers, including logistic regression classifier,random forest classifier, support vector classifier, and ensemble bagging, it wasobserved that the random forest classifier outperformed the other models in termsof predicting accuracy. This indicates that the random forest classifier was better atpredicting credit card approval.

Gesturing is a natural way of communication between people and is used in our everyday conversations. Hand gesture recognition systems are used in many applications in a wide variety of fields, such as mobile phone applications, smart TVs, video gaming, etc. With the advances in human-computer interaction technology, gesture recognition is becoming an active research area. There are two types of devices to detect gestures; contact based devices and contactless devices. Using ultrasonic waves for determining gestures is one of the ways that is employed in contactless devices. Hand gesture recognition utilizing ultrasonic waves will be the focus of this thesis work. This thesis presents a new method for detecting and classifying a predefined set of hand gestures using a single ultrasonic transmitter and a single ultrasonic receiver. This method uses a linear frequency modulated ultrasonic signal. The ultrasonic signal is designed to meet the project requirements such as the update rate, the range of detection, etc. Also, it needs to overcome hardware limitations such as the limited output power, transmitter, and receiver bandwidth, etc. The method can be adapted to other hardware setups. Gestures are identified based on two main features; range estimation of the moving hand and received signal strength (RSS). These two factors are estimated using two simple methods; channel impulse response (CIR) and cross correlation (CC) of the reflected ultrasonic signal from the gesturing hand. A customized simple hardware setup was used to classify a set of hand gestures with high accuracy. The detection and classification were done using methods of low computational cost. This makes the proposed method to have a great potential for the implementation in many devices including laptops and mobile phones. The predefined set of gestures can be used for many control applications.

No description available.

For most businesses there are costs involved when acquiring new customers and having longer relationships with customers is therefore often more profitable. Predicting if an individual is prone to leave the business is then a useful tool to help any company take actions to mitigate this cost. The event when a person ends their relationship with a business is called attrition or churn. Predicting peoples actions is however hard and many different factors can affect their choices. This paper investigates different machine learning methods for predicting attrition in the customer base of a bank. Four different methods are chosen based on the results they have shown in previous research and these are then tested and compared to find which works best for predicting these events. Four different datasets from two different products and with two different applications are created from real world data from a European bank. All methods are trained and tested on each dataset. The results of the tests are then evaluated and compared to find what works best. The methods found in previous research to most reliably achieve good results in predicting churn in banking customers are the Support Vector Machine, Neural Network, Balanced Random Forest, and the Weighted Random Forest. The results show that the Balanced Random Forest achieves the best results with an average AUC of 0.698 and an average F-score of 0.376. The accuracy and precision of the model are concluded to not be enough to make definite decisions but can be used with other factors such as profitability estimations to improve the effectiveness of any actions taken to prevent the negative effects of churn. / För de flesta företag finns det en kostnad involverad i att skaffa nya kunder. Längre relationer med kunder är därför ofta mer lönsamma. Att kunna förutsäga om en kund är nära att lämna företaget är därför ett användbart verktyg för att kunna utföra åtgärder för att minska denna kostnad. Händelsen när en kund avslutar sin relation med ett företag kallas här efter kundförlust. Att förutsäga människors handlingar är däremot svårt och många olika faktorer kan påverka deras val. Denna avhandling undersöker olika maskininlärningsmetoder för att förutsäga kundförluster hos en bank. Fyra metoder väljs baserat på tidigare forskning och dessa testas och jämförs sedan för att hitta vilken som fungerar bäst för att förutsäga dessa händelser. Fyra dataset från två olika produkter och med två olika användningsområden skapas från verklig data ifrån en Europeisk bank. Alla metoder tränas och testas på varje dataset. Resultaten från dessa test utvärderas och jämförs sedan för att få reda på vilken metod som fungerar bäst. Metoderna som enligt tidigare forskning ger de mest pålitliga och bästa resultaten för att förutsäga kundförluster hos banker är stödvektormaskin, neurala nätverk, balanserad slumpmässig skog och vägd slumpmässig skog. Resultatet av testerna visar att en balanserad slumpmässig skog får bäst resultat med en genomsnittlig AUC på 0.698 och ett F-värde på 0.376. Träffsäkerheten och det positiva prediktiva värdet på metoden är inte tillräckligt för att ta definitiva handlingar med men kan användas med andra faktorer så som lönsamhetsuträkningar för att förbättra effektiviteten av handlingar som tas för att minska de negativa effekterna av kundförluster.

Regular inspection and documentation for railway assets are necessary to monitor the status of the traffic environment. Mobile Laser Scanning (MLS) makes it possible to collect highly accurate spatial information of railway environments in the form of point cloud, and an automatic method to extract interested objects from the point cloud is needed to avoid too much manual work. In this project, point cloud along a railway in Saltsjöbanan was collected by MLS and processed to extract interested objects from it. The main purpose of the project is to develop a workflow for automatic extraction of masts of overhead supply and street lights from the study area. Researchers have proposed various methods for object extraction, such as model-based method, shape-based method, semantic method, and machine learning method recently. Different methods were reviewed and Support Vector Machine was chosen for the classification. Several softwares were reviewed as well. TerraScan and CloudCompare were chosen for pre-processing, and the major part was done in MATLAB. The proposed method consists of 4 steps: pre-processing, voxelization and segmentation, feature computation, classification and validation. The method calculates features to describe every object segmented from the point cloud and learns from the manually classified objects to train a classifier. The study area was divided into training data and validating data. The SVM classifier was trained using training data and evaluated using validating data. In the classification, 90.84% of the masts and 67.65% of the lights were correctly classified. There was some object loss during the step of pre-processing and segmentation. When including the loss from the pre-processing and segmentation step, 87.5% of the masts and 53.49% of the lights were successfully detected. The street lights have more various outlook and more complicated surrounding environment, which caused a relatively low accuracy. / Regelbunden inspektion och dokumentation för järnvägstillgångar är nödvändig för att övervaka trafikmiljön. Mobil Laser Scanning (MLS) gör det möjligt att samla in mycket exakt geografisk information om järnvägsmiljöer i form av punktmoln och en automatisk metod för att extrahera intresserade objekt från punktmoln är nödvändigt för att undvika för mycket manuellt arbete. I det här projektet samlades punktmoln längs en järnväg i Saltsjöbanan av MLS och bearbetades för att extrahera intresserade objekt från den. Huvudsyftet med projektet är att utveckla ett arbetsflöde för automatisk utvinning av kontaktledningsstolpar och gatubelysningsstolpar från studieområdet. Forskare har nyligen föreslagit olika metoder för objektutvinning som baseras på modell, form, semantisk och maskininlärning. I detta arbete har flera olika metoder för objektutvinning undersökts och slutligen valdes Support Vector Machine (SVM) för klassificering. Ett antal tillgängliga programvaror har utvärderats. TerraScan och CloudCompare valdes för förbehandling, och huvuddelen gjordes i MATLAB. Den föreslagna metoden består av 4 steg: förbehandling, voxelisering och segmentering, funktionen beräkning, klassificering och validering. Metoden beräknar funktioner för att beskriva varje objekt segmenterat från punktmoln och lär ut från de manuellt klassificerade objekten för att träna en klassificerare. Studieområdet delades in i träningsdata och validering av data. SVM-klassificeraren utbildades med träningsdata och utvärderades genom att validera data. I klassificeringen klassificerades 90,84% av kontaktledningsstolparna och 67,65% av belysningsstolparna korrekt. Det fanns vissa förluster av objekt under förbehandling och segmentering. Inkluderat förlusten i förbehandling och segmentering upptäcktes 87,5% av kontaktledningsstolparna och 53,49% av belysningsstolparna korrekt. Det något sämre resultatet vid detektion av belysningsstolpar beror på att dessa är placerade i en svårare miljö med närhet till andra objekt och inte minst vegitation. Att automatiskt detektera objekt i sådan miljö baserat på enbart laserdata är svårt vilket medförde en relativt låg noggrannhet.

The structure of regional correlation graphs built from fMRI-derived data is frequently used in algorithms to automatically classify brain data. Transformation on the data is performed during pre-processing to remove irrelevant or inaccurate information to ensure that an accurate representation of the subject's resting-state connectivity is attained. Our research suggests and confirms that such pre-processed data still exhibits inherent transitivity, which is expected to obscure the true relationships between regions. This obfuscation prevents known solutions from developing an accurate understanding of a subject’s functional connectivity. By removing correlative transitivity, connectivity between regions is made more specific and automated classification is expected to improve. The task of utilizing fMRI to automatically diagnose Attention Deficit/Hyperactivity Disorder was posed by the ADHD-200 Consortium in a competition to draw in researchers and new ideas from outside of the neuroimaging discipline. Researchers have since worked with the competition dataset to produce ever-increasing detection rates. Our approach was empirically tested with a known solution to this problem to compare processing of treated and untreated data, and the detection rates were shown to improve in all cases with a weighted average increase of 5.88%.

In this paper, we review existing classification techniques and suggest an entirely new procedure for the classification of high-dimensional vectors on the basis of a few training samples. The proposed method is based on the Bayesian paradigm and provides posterior probabilities that a new vector belongs to each of the classes, therefore it adapts naturally to any number of classes. Our classification technique is based on a small vector which is related to the projection of the observation onto the space spanned by the training samples. This is achieved by employing matrix-variate distributions in classification, which is an entirely new idea. In addition, our method mimics time-tested classification techniques based on the assumption of normally distributed samples. By assuming that the samples have a matrix-variate normal distribution, we are able to replace classification on the basis of a large covariance matrix with classification on the basis of a smaller matrix that describes the relationship of sample vectors to each other.