Incident detection is a key component in real-time traffic management systems that allows efficient response plan generation and decision making by means of risk alerts at critical affected sections in the network. State-of-the-art incident detection techniques traditionally require: i) good quality data from closely located sensor pairs, ii) a minimum of two reliable measurements from the flow- occupancy-speed triad, and iii) supervised adjustment of thresholds that will trigger anomalous traffic states. Despite such requirements may be reasonably achieved in simulated scenarios, real-time downstream applications rarely work under such ideal conditions and must deal with low reliability data, missing measurements, and scarcity of curated incident labelled datasets, among other challenges. This paper proposes an unsupervised technique based on univariate timeseries anomaly detection for computationally efficient incident detection in real-world scenarios. Such technique is proved to successfully work when only flow measurements are available, and to dynamically adjust thresholds that adapt to changes in the supply. Moreover, results show good performance with low-reliability and missing data.

People analytics (PA) has experienced significant growth in recent years due to the increasing availability of employee data and the impact of digitalization on organizations. This data-driven approach utilizes inductive methods to predict various outcomes in the field of human resources. Nevertheless, concerns have emerged regarding the availability and reliability of the data used in PA. Surprisingly, the quality standards of these data-driven methods have not been evaluated in the PA literature, despite their widespread adoption. To address these gaps, nine research questions covering expertise areas, psychological constructs, patterns/trends, study types, data sources, reliability reporting, data-driven frameworks, prediction accuracy, and open science practices in PA were reviewed. A scoping review was conducted to extract relevant information from each piece of literature, while bibliometric analysis provides a structured analysis of trends, themes, and key contributors. A total of 3,103 records were identified from the Scopus (n = 449) and APA PsycINFO (n = 2,700) databases, with nine studies included in the review. Findings indicated a lack of consideration given to quality, reliability aspects, and open science practices within PA literature. The predominant emphasis of the research was on the evaluation of variables, particularly turnover intention. This study contributes to advancing the understanding of PA by emphasizing the importance of incorporating quality standards and open science practices to enhance the reliability and credibility of research findings. The classification of the PA literature and recommendations for future research directions are provided, highlighting the need for a hierarchy of knowledge in the field. / Scoping Review of People Analytics

Organizations are social systems established to make decisions. Modern organizational decision-making is complex and can easily overwhelm the capacity of individuals. Because of the complexity of multi-person decisions, there is a big risk for uncertainty in decision-making. In recent years, the rise of business intelligence has enabled organizations to base their decisions on data and minimize uncertainty in their decision-making. However, deployment of business intelligence systems is characterized by complexity, making many small and medium-sized organizations fail to use such a system effectively.This thesis aims to identify and describe variables that influence successful use of a business intelligence architecture to support small organizations in making data-based decisions, what small organizations need to become data-driven in decision-making, and what measures small organizations can take to use business intelligence systems efficiently. Eight semi-structured interviews were conducted with professionals from Vasaloppet, a small organization deploying a business intelligence system. The empirical data gathered have been analyzed with a thematic approach. The thematic analysis identified four themes’ Deficiencies in organizational governance, Deficiencies in data management, Perceived workload, and Degree of matching between processes, organization, and strategy. Findings in these themes and underlying codes within these themes revealed problem areas in organizational governance when making decisions. Respondents mentioned challenges with a lack of a decision model, clear business plan, and intra-organizational understanding. When it comes to becoming data-driven, respondents said deficiency of structure for communication, lack of access to data, lack of data in decision-making, general workload, deficiencies in project results, and deficiencies in degree of matching as problematic. Based on the results of this study, guidelines are presented for small organizations to become data-driven in their decision-making.Keywords: Data-driven decision-making, business intelligence, small organizations

Indiana University-Purdue University Indianapolis (IUPUI)

Manufacturers can develop next-generation production and service for their customers by the data gathered and analyzed from customers’ usage conditions. In this research, the operating condition of articular haulers is collected and analyzed through machine learning algorithms to predict the type of operational topographies and road surface. To achieve that, elevation data and satellite images, which were gathered from Microsoft Azure Maps, are used as data sources to identify the topography and road surface on which machines operated. In the end, two machine learning models are trained with machines’ inclination records and road roughness records, respectively, to classify the topography and road surface. For the topography classifier, the topography is categorized into four terrain labels, including "Low Hills", "Mountains", "Plains", and "Tablelands & High Hills". The road surface is classified into "Paved" and "Unpaved". A Convolutional Neural Network (CNN) image classification model is built for labeling satellite images instead of labeling manually. The results indicate that the prediction for topography labels "Plains" and "Tablelands & High Hills" has superior performance, which accounts for the majority of the raw dataset; on the contrary, the road surface classifier still needs further improvement in the future. In addition, an analysis and discussion regarding the imbalanced dataset are included, and it shows the limited effect on an extremely imbalanced dataset. Finally, the conclusion and future work are given.

During the past years an increased focus on the development of autonomous solutions has resulted in driverless vehicles being used in numerous industries. Volvo Construction Equipment is currently developing the TA15, an autonomous hauler part of a larger transport solution. The transition to autonomous haulers have further increased the need for improved system condition monitoring in the strive for increased operational time. A method aiming to identify and isolate faults in the hydraulic steering system on the TA15 was therefore investigated in this thesis. Using fault tree analysis, five faults considered to be of importance regarding steering performance were selected. Two different methods for detecting the faults were compared to each other, data-driven and model based. Out of the two, data-driven was selected as the method of choice due to high modularity and relative simplicity regarding implementation. The data-driven approach consisted of Feed-Forward and Long Short Term Memory networks where the suitable inputs were decided to be a combination of pressure and position signals. Utilizing a simulation model of the steering system validated against the TA15, the selected faults were induced in the simulated system with various severity. Training the networks to classify and estimate fault severity in the simulated model resulted in satisfactory results using both networks. It was however concluded that in contrary to the Feed-Forward network, the LSTM network could achieve good performance using less amount of sensors. Although the diagnostic method showed promising result on a simulation model, test on the real TA15 needs to be performed in order to properly evaluate the method. The advantage of using a data-driven approach was specially noticeable when comparisons were made to the model based approach. The data-driven approach relies on labeling data rather than complete system knowledge. Meaning that the method developed therefore could be applied on practically any hydraulic system in construction equipment by changing the training data.

The term electronic commerce (e-commerce) refers to a business model that allows companies and individuals to buy and sell goods and services over the internet. The focus of this thesis is on online pharmacies, a segment of the ecommerce market. Even though internet pharmacies are still subject to the same stringent rules imposed on pharmacies that limit the scope for their market growth, it has shown a notable increase in the past decades. The main goal of this thesis is to develop a data-driven marketing strategy based on a Swedish based online pharmacy’s daily sales data. The methodology of the data analysis includes exploratory data analysis (EDA) and market basket analysis (MBA) using the Apriori algorithm and the application of marketing frameworks and theories from a data-driven standpoint. In addition to the data analysis, this paper proposes a conceptual framework of a digital marketing strategy based on the RACE framework (reach, act, convert, and engage). The result of the analysis has led to the following data-driven marketing strategy: Special attention should be paid to association rules with a high lift ration value; high gross profit margin percentile (GPMP) products should have a volume-based marketing strategy that focuses on lower prices on subsequent items; and price bundling is the best marketing strategy for low GPMP products. Some of the practical ideas mentioned in this thesis paper include optimizing keyword search for a high GPMP product type and sending reminder emails and push alerts to avoid cart abandonment. The findings and recommendations presented in this thesis can be used by online pharmacies to extract knowledge that may support several decisions ranging from raising overall order size, marketing campaigns, to increasing the sales of products with a high gross profit margin.

In this report, problems regarding the retrieving, measuring, and analysis of data when analysing marketing effects in the web analytics tool Google Analytics will be discussed. A correct setup, configuration, maintenance, campaign tracking and the understanding of the data in Google Analytics is essential to be able to achieve relevant insights. This is important since many Swedish marketing departments experience issues related to their setup of Google Analytics as well as the ongoing configuration and maintenance. A literature study has been conducted to gather information, focusing on collecting theories from researchers and experts in the field of web analytics and marketing analytics. Google Analytics data and reports from several Swedish companies have been studied to gain a deep understanding of how the tool is used for the measuring and analysis of the marketing effects. Interviews with marketing department and media bureau/agency employees have been conducted and analysed in a qualitative manner. A thematic analysis of the interviews has been done, resulting in 8 themes which are presented in the result section. The result has been analysed and discussed in relation to the theory. The interviews showed that there is a difference in knowledge and experience between the senior and junior analysts, and that there is a significant learning curve when working in Google Analytics. The junior analysts trusted the data, and did not know about campaign tracking and filters, in contrast to the senior analysts, who did not trust the data as a control mechanism, and did work with campaign tracking and filters. Furthermore, the senior analysts had more understanding of the data models in Google Analytics, such as attribution models, which are known to show different stories based on which attribution model is being used. The conclusions are four capabilities that address a need for more and better control over the setup and over the data, a wider use of campaign tracking, and wider knowledge of the data and the data models in Google Analytics, and of the business the organisation is conducting, to be able to gain relevant insights. / I den här rapporten diskuteras problemen med att insamla, mäta och analysera data vid analys av marknadseffekter i webbanalys-verktyget Google Analytics. Korrekt installation, konfiguration, underhåll, kampanjspårning och förståelsen av datan i Google Analytics är viktigt för att kunna uppnå relevanta insikter. Detta är viktigt eftersom att många svenska marknadsföringsavdelningar upplever problem i samband med installationen av Google Analytics samt den pågående konfigurationen och underhållet av data som ska mätas och analyseras. En litteraturstudie har gjorts för att samla in information, med inriktning på att samla teori från forskare och experter inom webbanalys och marknadsanalys. Google Analytics-data och rapporter från flera svenska företag har studerats för att få en djupare förståelse för hur verktyget används för att mäta och analysera marknadsföringseffekter. Intervjuer med medarbetare på marknadsavdelningar och mediebyråer har genomförts och analyserats på ett kvalitativt sätt. En tematisk analys av intervjuerna har gjorts, vilket resulterat i 8 teman som presenteras i resultatavsnittet. Resultatet har analyserats och diskuterats i förhållande till teorin. Intervjuerna visade att det finns skillnad i kunskap och erfarenhet mellan seniora och juniora analytiker, och att det finns en signifikant inlärningskurva när en arbetar i Google Analytics. De juniora analytikerna litade på datan och tillämpade inte kampanjspårning och filter i motsats till de seniora analytikerna som inte litade på datan som en kontrollmekanism, samt tillämpade kampanjspårning och filter. Vidare hade de seniora analytikerna större förståelse för datamodellerna i Google Analytics, till exempel attributionsmodeller, som är kända för att indikera olika saker baserat på vilken modell som används. Slutsatserna är fyra förmågor som relaterar till ett behov av mer och bättre kontroll över datan och installationen av Google Analytics, en bredare användning av kampajspårning, bredare kunskaper om både datan och de olika datamodellerna i Google Analytics, och verksamheten som organisationen utför för att kunna tillskansa sig relevanta insikter som är lämpliga att grunda beslut utifrån.

Stress and overload are strongly associated with unsafe behaviour, which motivated various studies to detect them automatically in workplaces. This study aims to advance safety research by developing a data-driven stress and overload detection method. An unsupervised deep learning-based anomaly detection method is developed to detect stress. The proposed method performs with convolutional neural network encoder-decoder and long short-term memory equipped with an attention layer. Data from a field experiment with 18 participants was used to train and test the developed method. The field experiment was designed to include a pre-defined sequence of activities triggering mental and physical stress, while a wristband biosensor was used to collect physiological signals. The collected contextual and physiological data were pre-processed and then resampled into correlation matrices of 14 features. Correlation matrices are used as an input to the unsupervised Deep Learning (DL) based anomaly detection method. The developed method is validated, offering accuracy and F-measures close to 0.98. The technique employed captures the input data attributes correlation, promoting higher interpretability of the DL method for easier comprehension. Over-reliance on uncertain absolute truth, the need for a high number of training samples, and the requirement of a threshold for detecting anomalies are identified as shortcomings of the proposed method. To overcome these shortcomings, an Adaptive Neuro-Fuzzy Inference System (ANFIS) was designed and developed. While the ANFIS method did not improve the overall accuracy, it outperformed the DL-based method in detecting anomalies precisely. The overall performance of the ANFIS method is better than the DL-based method for the anomalous class, and the method results in lower false alarms. However, the DL-based method is suitable for circumstances where false alarms are tolerated. / Dissertation / Doctor of Philosophy (PhD)

No description available.