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 Bag of Little Bootstraps (BLB) was introduced to make the bootstrap method more computationally efficient when used on massive data samples. Since its introduction, a broad spectrum of research on the application of the BLB has been made. However, while the BLB has shown promising results that can be used for logistic regression, these results have been for well-balanced data. There is, therefore, an obvious need for further research into how the BLB performs when the dependent variable is unbalanced and whether possible performance issues can be remedied through methods such as Firths's Penalized Maximum Likelihood Estimation (PMLE). This thesis shows that the dependent variable's imbalances severely affect the BLB's performance when applied in logistic regression. Further, this thesis also shows that PMLE produces mixed and unreliable results when used to remedy the drops in performance.

Explainability is a key component in building trust for computer calculated predictions when they are applied to areas with influence over individual people. This bachelor thesis project report focuses on the explanation regarding the decision making process of the machine learning method Logistic Regression when predicting mortality. The aim is to present theoretical information about the predictive model as well as an explainable interpretation when applied on the clinical MIMIC-III database. The project found that there was a significant difference between particular features considering the impact of each individual feature on the classification. The feature that showed the greatest impact was the Glasgow Coma Scale value, which could be proven through the fact that a good classifier could be constructed with only that and one other feature. An important conclusion from this study is that a great focus should be enforced early in the implementation process when the features are selected. In this specific case, when medical artificial intelligence is implemented, medical expertise is desired in order to make a good feature selection. / Förklarbarhet är en viktig komponent för att skapa förtroende för datorframtagna prognoser när de appliceras på områden som påverkar individuella personer. Denna kandidatexamensarbetesrapport fokuserar på förklarandet av beslutsprocessen hos maskininlärningsmetoden Logistic Regression när dödlighet ska förutsägas. Målet är att presentera information om den förutsägande modellen samt en förklarbar tolkning av resultaten när modellen appliceras på den kliniska databasen MIMIC-III. Projektet fann att det fanns signifikanta skillnader mellan särskilda egenskaper med hänsyn till den påverkan varje enskild egenskap har på klassificeringen. Den egenskapen som visade ha störst inverkan var Glascow Coma Scale värdet, vilket kunde visas via det faktum att en god klassificerare kunde konstrueras med endast den och en annan egenskap. En viktig slutsats av denna studie är att stort fokus bör läggas tidigt i implementationsprocessen då egenskaperna väljs. I detta specifika fall, då medicinsk artificiell intelligens implementeras, krävs medicinsk expertis för att göra ett gott egenskapsurval. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm

The dissertation focuses on the different determinants of education, their effects on the educational outcome, and the overall effect of education on the lifetime consequences. The first chapter focuses on the inequality of educational opportunity across different demographic factors. This chapter employs a broader set of social factors to provide fresh insights into the inequality situation in the USA relative to those of the extant literature. The chapter employs polynomial trends for the effects of social factors to identify long-term trends in the determinants of the differences in attainment of each of four achievements (high school graduation, some college, college graduation, and post-college work) across different endogenous social groups. Using the Panel Study of Income Dynamics (PSID) data for the years of 1968-2013, we show how inequality of educational opportunity and its determinants have evolved over the years. The chapter utilizes the machine-learning process and logistic regression model to identify inequality of opportunity. The second chapter examines the age demographic distribution of graduates across cohorts from 1940 until 1990. Using the PSID data, the paper explored the first and second moment of the age of graduating from high school and college across the US. To deal with the data deficiencies, a large part of the chapter dealt with data preparation. The chapter provides a unique method of extracting information on the graduating age of the individuals both from high school and from college. The results show a large dispersion across the full sample. The data truncated to a standard length, however, provides a much smaller dispersion and much smaller moments. The chapter concludes that as the time passes, people tend to attain education at a younger age. The third chapter investigates the trends of the contribution of different factors of income starting from 1910 cohort. Following Mincer (1974), a wave of papers studied how various factors contribute to the earnings of individuals. This paper contributes to that literature in three ways: (i) using the PSID data, it computes the actual working experience of the individuals, (ii) it studies the cohorts who were born in 1910 or afterwards, unlike the existing papers, and (iii) it adds two variables—technological progress and the occupation with which individuals start their careers—to an extended Mincerian equation. The results re-emphasize the importance of education in lifetime earnings. The results also show that while some of the determinants of income have become more important over the years, other factors have not changed much in importance. / PHD / The reason for choosing the theme ‘Evolution of the Determinants of Educational Attainment and its Consequences’ was to investigate the different determinants of education, their effects on the educational outcome, and the overall effect of education on the lifetime consequences. Education is considered as one of the tools to eradicate poverty. Yet, countries with high educational coverage keeps suffering from poverty, a reason for which is higher inequality of opportunity. In the first chapter, entitled ‘Inequality in Educational Opportunity in the United States’, opportunity inequality in education is illustrated. Much inequality stems from differences in educational attainment. A lack of educational attainment puts an individual behind in the career race, even before the race has started. While individuals are responsible for some of the differences in educational attainment, there are factors outside the control of individuals that play substantial roles. The inequality that arises from these factors is known as inequality of opportunity. This paper focuses on inequality of educational opportunity across socioeconomic background, race, and sex. The factors that are analyzed for their contributions to inequality of educational opportunity are father’s education, father’s occupation, mother’s education, and economic status of the individual’s family. The results show that inequality of opportunity has seen a consistent decline for high school completion. The inequality of opportunity (IO) declines for obtaining some college education for the bottom two social groups and remained persistent for the relatively more advantaged group. For college/post-college education, the IO is much lower and, in general, remained persistent across the social strata. Although the females were behind the males – given the equal opportunity – regardless of the race and socioeconomic status during the beginning and the mid twentieth century, the scenario reversed in the late twentieth century. In terms of educational disparity among races, African Americans trail their White counterparts along all the years. The second chapter ‘First and Second Moments of the Age Distributions of Graduates’ looks into the age characteristics (mean and variance) in graduating from high school and college across the cohorts from 1940s to 1990s. The idea of the paper largely came from the first chapter of the dissertation as we assumed the lack of opportunity at the earlier age could delay the attainment of education. The paper intends to find out the average age of graduation over the years. In the process, the paper put forward a method to extract the information of age of graduation from the Panel Study of Income Dynamics (PSID) data, as the database does not readily avail the information. The chapter concludes that as the time passes, people tend to attain education at a much younger age. Titled as ‘Factors Affecting Income: Education, Experience, and Beyond’, the third chapter investigates the contribution of different factors – education, experience, parental endowments, and labor market conditions – in the returns to education using the PSID data and compare the more recent scenarios with the past. This paper focuses on the trend of the rate of return to different factors of income across the two cohorts – those born between 1910 and 1950, and those born after 1950 – while identifying the changes in the returns for the same education level over time. The paper aims to find out how the contribution of the different factors of earning has changed in the USA over the years. The paper also intends to find out the role of technological progress in reducing the earning gaps across the different social groups. The results re-emphasize the importance of education in lifetime earnings. Experience has become a more important factor of income over the years. The chapter also suggests that income of an individual is a monotonic function of socioeconomic endowments and better endowments resulted in higher returns. Lastly, the chapter finds that the technological investment is progressive in manner.

As the use of AI in society is developing, the requirement of explainable algorithms has increased. A challenge with many modern machine learning algorithms is that they, due to their often complex structures, lack the ability to produce human-interpretable explanations. Research within explainable AI has resulted in methods that can be applied on top of non- interpretable models to motivate their decision bases. The aim of this thesis is to compare an unexplained machine learning model used in combination with an explanatory method, and a model that is explainable through its inherent structure. Random forest was the unexplained model in question and the explanatory method was SHAP. The explainable model was logistic regression, which is explanatory through its feature weights. The comparison was conducted within the area of creditworthiness and was based on predictive performance and explainability. Furthermore, the thesis intends to use these models to investigate what characterizes loan applicants who are likely to default. The comparison showed that no model performed significantly better than the other in terms of predictive performance. Characteristics of bad loan applicants differed between the two algorithms. Three important aspects were the applicant’s age, where they lived and whether they had a residential phone. Regarding explainability, several advantages with SHAP were observed. With SHAP, explanations on both a local and a global level can be produced. Also, SHAP offers a way to take advantage of the high performance in many modern machine learning algorithms, and at the same time fulfil today’s increased requirement of transparency. / I takt med att AI används allt oftare för att fatta beslut i samhället, har kravet på förklarbarhet ökat. En utmaning med flera moderna maskininlärningsmodeller är att de, på grund av sina komplexa strukturer, sällan ger tillgång till mänskligt förståeliga motiveringar. Forskning inom förklarar AI har lett fram till metoder som kan appliceras ovanpå icke- förklarbara modeller för att tolka deras beslutsgrunder. Det här arbetet syftar till att jämföra en icke- förklarbar maskininlärningsmodell i kombination med en förklaringsmetod, och en modell som är förklarbar genom sin struktur. Den icke- förklarbara modellen var random forest och förklaringsmetoden som användes var SHAP. Den förklarbara modellen var logistisk regression, som är förklarande genom sina vikter. Jämförelsen utfördes inom området kreditvärdighet och grundades i prediktiv prestanda och förklarbarhet. Vidare användes dessa modeller för att undersöka vilka egenskaper som var kännetecknande för låntagare som inte förväntades kunna betala tillbaka sitt lån. Jämförelsen visade att ingen av de båda metoderna presterande signifikant mycket bättre än den andra sett till prediktiv prestanda. Kännetecknande särdrag för dåliga låntagare skiljde sig åt mellan metoderna. Tre viktiga aspekter var låntagarens °ålder, vart denna bodde och huruvida personen ägde en hemtelefon. Gällande förklarbarheten framträdde flera fördelar med SHAP, däribland möjligheten att kunna producera både lokala och globala förklaringar. Vidare konstaterades att SHAP gör det möjligt att dra fördel av den höga prestandan som många moderna maskininlärningsmetoder uppvisar och samtidigt uppfylla dagens ökade krav på transparens.

Företag inom dagligvaruhandeln använder sig ofta av database marketing för att anpassa deras erbjudande till deras kunder och därmed stärka kundrelationen och ökaderas försäljning. Länge har logistisk regression varit en modell som ofta används för att bygga upp maskininlärningsmodeller som kan förutse vilka erbjudanden som löses in av vilken kund. I arbetet utvärderas en maskininlärningsmodell med logistisk regression och stepwise selection på kunddata från en av Sveriges större aktörer inom dagligvaruhandeln. Modellen jämförs med en annan modell som istället använder sig utav elastic net, vilket är en regulariserad regressionsmetod. Modellerna testas på fem olika produkter ur företagets sortiment och baseras på ett femtiotal variabler som beskriver kundernas sociodemografiska data och historiska köpbeteende i företagets butiker. Dessa utvärderas med hjälp av en förväxlingsmatris och värden för deras Accuracy, Balanced Accuracy, Precision, Recall och F1-score. Dessutom utvärderas modellen utifrån affärsnytta, påverkan på kundrelationer och hållbarhet. Studien visade att den logistiska regressionen med stepwise selection hade ett genomsnittligt värde för Precision på 23 procent. Vid användning av elastic net ökade värdet för Precision med i genomsnitt 7 procentenheter för samtliga modeller. Detta kan bero på att vissa av parametrarna i modellen med stepwise selection får överdrivet stora värden samt att stepwise selection väljer ut variabler för modellen som inte är optimala för att förutsäga kundens beteende. Det noterades även att kunder generellt verkade nöjda med de erbjudanden de fått, men missnöjda ifall de kände sig missförstådda av företaget. / Companies within the FMCG trade often uses database marketing to customize offers to each customer, and thereby strengthen customer relationships to the company and increase their sales. For a long time, logistic regression has been the preferred machine modelling method to predict which offer to present to each costumer. This study evaluates a machinelearning model based on logistic regression and stepwise selection on costumer data from one of Sweden’s larger companies within the FMCG trade. The model is later compared to another model based on the elastic net-method, which is a regularized regressionmodel. The models are tested on five different products from the company’s assortment and are based on about fifty different variables which describes the costumers’ sociodemographic factors and purchasing history. The models are evaluated using a confusion matrix and values stating their Accuracy, BalancedAccuracy, Precision, Recall and F1-score. Furthermore, the model is evaluated in the perspectives of business advantages, costumer relations and sustainability. The study concluded that the logistic regression and stepwise selection-model had an average Precisionon 23 procent. When the elastic net-method was used the Precision increased with approximately 7 percentage points. This might depend on the fact that some of the parameters in the logistic regression-model had an overrated value and that the stepwise selection chose a subset of features that was not optimal to predict the consumer behaviour. It was also noted that costumers most often seemed content, but were dissatisfied if they felt misunderstood by the company.

Numerous invertebrates have contributed to our understanding of the biology of learning and memory. In most cases, learning performance is documented for groups of individuals, and nearly always based on a single, typically binary, behavioural metric for a conditioned response. This is unfortunate for several reasons. Foremost, it has become increasingly apparent that invertebrates exhibit inter-individual differences in many aspects of their behaviour, and also that the conditioned response probability for an animal group does not adequately represent the behaviour of individuals in classical conditioning. Furthermore, a binary response character cannot yield a graded score for each individual. We also hypothesise that due to the complexity of a conditioned response, a single metric need not reveal an individual’s full learning potential. In this paper, we report individual learning scores for freely moving adult male crickets (Gryllus bimaculatus) based on a multi-factorial analysis of a conditioned response. First, in an absolute conditioning paradigm, we video-tracked the odour responses of animals that, in previous training, received either odour plus reward (sugar water), reward alone, or odour alone to identify behavioural predictors of a conditioned response. Measures of these predictors were then analysed using binary regression analysis to construct a variety of mathematical models that give a probability for each individual that it exhibited a conditioned response (Presp). Using standard procedures to compare model accuracy, we identified the strongest model which could reliably discriminate between the different odour responses. Finally, in a differential appetitive olfactory paradigm, we employed the model after training to calculate the Presp of animals to a conditioned, and to an unconditioned odour, and from the difference a learning index for each animal. Comparing the results from our multi-factor model with a single metric analysis (head bobbing in response to a conditioned odour), revealed advantageous aspects of the model. A broad distribution of model-learning scores, with modes at low and high values, support the notion of a high degree of variation in learning capacity, which we discuss.

No description available.

No description available.

No description available.