A data driven approach for automating vehicle activated signs

Vehicle activated signs (VAS) display a warning message when drivers exceed a particular threshold. VAS are often installed on local roads to display a warning message depending on the speed of the approaching vehicles. VAS are usually powered by electricity; however, battery and solar powered VAS are also commonplace. This thesis investigated devel-opment of an automatic trigger speed of vehicle activated signs in order to influence driver behaviour, the effect of which has been measured in terms of reduced mean speed and low standard deviation. A comprehen-sive understanding of the effectiveness of the trigger speed of the VAS on driver behaviour was established by systematically collecting data. Specif-ically, data on time of day, speed, length and direction of the vehicle have been collected for the purpose, using Doppler radar installed at the road. A data driven calibration method for the radar used in the experiment has also been developed and evaluated. Results indicate that trigger speed of the VAS had variable effect on driv-ers’ speed at different sites and at different times of the day. It is evident that the optimal trigger speed should be set near the 85th percentile speed, to be able to lower the standard deviation. In the case of battery and solar powered VAS, trigger speeds between the 50th and 85th per-centile offered the best compromise between safety and power consump-tion. Results also indicate that different classes of vehicles report differ-ences in mean speed and standard deviation; on a highway, the mean speed of cars differs slightly from the mean speed of trucks, whereas a significant difference was observed between the classes of vehicles on lo-cal roads. A differential trigger speed was therefore investigated for the sake of completion. A data driven approach using Random forest was found to be appropriate in predicting trigger speeds respective to types of vehicles and traffic conditions. The fact that the predicted trigger speed was found to be consistently around the 85th percentile speed justifies the choice of the automatic model.

Ensidesapplikationers kodkvalitet och förvaltningsbarhet : En jämförelse av förvaltningsbarhet hos single-page applications utvecklade med AngularJS och React

Single-page applications have historically been subject to strong market forces driving fast development and deployment in lieu of quality control and changeable code, which are important factors for maintainability. In this report we develop two functionally equivalent applications using AngularJS and React and compare their maintainability as defined by ISO/IEC 9126. AngularJS and React represent two distinct approaches to web development, with AngularJS being a general framework providing rich base functionality and React a small specialized library for efficient view rendering. The quality comparison was accomplished by calculating Maintainability Index for each application. Version control analysis was used to determine quality indicators during development and subsequent maintenance where new functionality was added in two steps. The results show no major differences in maintainability in the initial applications. As more functionality is added the Maintainability Index decreases faster in the AngularJS application, indicating a steeper increase in complexity compared to the React application. Source code analysis reveals that changes in data flow requires significantly larger modifications of the AngularJS application due to its inherent architecture for data flow. We conclude that frameworks are useful when they facilitate development of known requirements but less so when applications and systems grow in size. Sammanfattning: Ensidesapplikationer har historiskt sett påverkats av starka marknadskrafter som pådriver snabba utvecklingscykler och leveranser. Detta medför att kvalitetskontroll och förändringsbar kod, som är viktiga faktorer för förvaltningsbarhet, blir lidande. I denna rapport utvecklar vi två funktionellt ekvi-valenta ensidesapplikationer med AngularJS och React samt jämför dessa applikationers förvaltningsbarhet enligt ISO/IEC 9126. AngularJS och React representerar två distinkta angreppsätt på webbutveckling, där AngularJS är ett ramverk med mycket färdig funktionalitet och React ett mindre bibliotek specialiserat på vyrendering. Kvalitetsjämförelsen utfördes genom att beräkna förvaltningsbarhetsindex för respektive applikation. Versionshanteringsanalys användes för att bestämma andra kvalitetsindikatorer efter den initiala utvecklingen samt två efterföljande underhållsarbeten. Resultaten visar inga markanta skillnader i förvaltningsbarhet för de initiala applikationerna. I takt med att mer funktionalitet lades till sjönk förvaltnings-barhetsindex snabbare för AngularJS-applikationen, vilket motsvarar en kraftigare ökning i komplexitet jämfört med React-applikationen. Versionshanteringsanalys visar att ändringar i dataflödet kräver större modifikationer för AngularJS-applikationen på grund av dess förbestämda arkitektur. Utifrån detta drar vi slutsatsen att ramverk är användbara när de understödjer utvecklingen mot kända krav men att deras nytta blir begränsad ju mer en applikation växer i storlek.

LDA-TM : A two-step approach to twitter topic data clustering

The Twitter System is the biggest social network in the world, and everyday millions of tweets are posted and talked about, expressing various views and opinions. A large variety of research activities have been conducted to study how the opinions can be clustered and analyzed, so that some tendencies can be uncovered. Due to the inherent weaknesses of the tweets - very short texts and very informal styles of writing - it is rather hard to make an investigation of tweet data analysis giving results with good performance and accuracy. In this paper, we intend to attack the problem from another aspect - using a two-layer structure to analyze the twitter data: LDA with topic map modelling. The experimental results demonstrate that this approach shows a progress in twitter data analysis. However, more experiments with this method are expected in order to ensure that the accurate analytic results can be maintained.

Collaborative Filtering with Data Classification : A Combined Approach to Hotel Recommendation Systems

Recommendation systems aim to help users make decisions more efficiently. The most widely used method in recommendation systems is collaborative filtering, of which, a critical step is to analyze a user's preferences and make recommendations of products or services based on similarity analysis with other users' ratings. However, collaborative filtering is less usable for recommendation facing the "cold start" problem, i.e. few comments being given to products or services. To tackle this problem, we propose an improved method that combines collaborative filtering and data classification. We use hotel recommendation data to test the proposed method. The accuracy of the recommendation is determined by the rankings. Evaluations regarding the accuracies of Top-3 and Top-10 recommendation lists using the 10-fold cross-validation method and ROC curves are conducted. The results show that the Top-3 hotel recommendation list proposed by the combined method has the superiority of the recommendation performance than the Top-10 list under the cold start condition in most of the times.

Perspektiv på språkprocessering

Automatisk språkprocessering har efter mer än ett halvt sekel av forskning blivit ett mycket viktigt område inom datavetenskapen. Flera vetenskapligt viktiga problem har lösts och praktiska applikationer har nått programvarumarknaden. Disambiguering av ord innebär att hitta rätt betydelse för ett mångtydigt ord. Sammanhanget, de omkringliggande orden och kunskap om ämnesområdet är faktorer som kan användas för att disambiguera ett ord. Automatisk sammanfattning innebär att förkorta en text utan att den relevanta informationen går förlorad. Relevanta meningar kan plockas ur texten, eller så kan en ny, kortare text genereras på basen av fakta i den ursprungliga texten. Avhandlingen ger en allmän översikt och kort historik av språkprocesseringen och jämför några metoder för disambiguering av ord och automatisk sammanfattning. Problemområdenas likheter och skillnader lyfts fram och metodernas ställning inom datavetenskapen belyses.

From the road network database to a graph for localization purposes

The problems of finding best facility locations require complete and accurate road network with the corresponding population data in a specific area. However the data obtained in road network databases usually do not fit in this usage. In this paper we propose our procedure of converting the road network database to a road graph which could be used in localization problems. The road network data come from the National road data base in Sweden. The graph derived is cleaned, and reduced to a suitable level for localization problems. The population points are also processed in ordered to match with that graph. The reduction of the graph is done maintaining most of the accuracy for distance measures in the network.

A web application for follow-up of results from a mobile device test battery for parkinson’s disease patients

Background: A test battery consisting of self-assessments and motor tests (tapping and spiral drawing) was developed for a hand computer with touch screen in a telemedicine setting. Objectives: To develop and evaluate a web-based system that delivers decision support information to the treating clinical staff for assessing PD symptoms in their patients based on the test battery data. Methods: The test battery is currently being used in a clinical trial (DAPHNE, EudraCT No. 2005-002654-21) by sixty five patients with advanced Parkinson’s disease (PD) on 9991 test occasions (four tests per day during in all 362 week-long test periods) at nine clinics around Sweden. Test results are sent continuously from the hand unit over a mobile net to a central computer and processed with statistical methods. They are summarized into scores for different dimensions of the symptom state and an ‘overall test score’ reflecting the overall condition of the patient during a test period. The information in the web application is organized and presented graphically in a way that the general overview of the patient performance per test period is emphasized. Focus is on the overall test score, symptom dimensions and daily summaries. In a recent preliminary user evaluation, the web application was demonstrated to the fifteen study nurses who had used the test battery in the clinical trial. At least one patient per clinic was shown. Results: In general, the responses from nurses were positive. They claimed that the test results shown in the system were consistent with their own clinical observations. They could follow complications, changes and trends within their patients. Discussion: In conclusion, the system is able to summarise the various time series of motor test results and self-assessments during test periods and present them in a useful manner. Its main contribution is a novel and reliable way to capture and easily access symptom information from patients’ home environment. The convenient access to current symptom profile as well as symptom history provides a basis for individualized evaluation and adjustment of treatments.

Spiral drawing during self-rated dyskinesia is more impaired than during self-rated off

Background: A mobile device test battery, consisting of a patient diary collection section with disease-related questions and a fine motor test section (including spiral drawing tasks), was used by 65 patients with advanced Parkinson's disease (PD)(treated with intraduodenal levodopa/carbidopa gel infusion, Duodopa®, or candidates for this treatment) on 10439 test occasions in their home environments. On each occasion, patients traced three pre-drawn Archimedes spirals using an ergonomic stylus and self-assessed their motor function on a global Treatment Response Scale (TRS) ranging from -3 = very 'off' to 0 = 'on' to +3 = very dyskinetic. The spirals were processed by a computer-based method that generates a "spiral score" representing the PD-related drawing impairment. The scale for the score was based on a modified Bain & Findley rating scale in the range from 0 = no impairment to 5 = moderate impairment to 10 = extremely severe impairment. Objective: To analyze the test battery data for the purpose to find differences in spiral drawing performance of PD patients in relation to their self-assessments of motor function. Methods: Three motor states were used in the analysis; OFF state (including moderate and very 'off'), ON state ('on') and a dyskinetic (DYS) state (moderate and very dyskinetic). In order to avoid the problem of multiple test occasions per patient, 200 random samples of single test occasions per patient were drawn. One-way analysis of variance, ANOVA, test followed by Tukey multiple comparisons test was used to test if mean values of spiral test parameters, i.e. the spiral score and drawing completion times (in seconds), were different among the three motor states. Statistical significance was set at p<0.05. To investigate changes in the spiral score over the time-of-day test sessions for the three motor states, plots of statistical summaries were inspected. Results: The mean spiral score differed significantly across the three self-assessed motor states (p<0.001, ANOVA test). Tukey post-hoc comparisons indicate that the mean spiral score (mean ± SD; [95% CI for mean]) in DYS state (5.2 ± 1.8; [5.12, 5.28]) was higher than the mean spiral score in OFF (4.3 ± 1.7; [4.22, 4.37]) and ON (4.2 ± 1.7; [4.17, 4.29]) states. The mean spiral score was also significantly different among individual TRS values of slightly 'off' (4.02 ± 1.63), 'on' (4.07 ± 1.65) and slightly dyskinetic (4.6 ± 1.71), (p<0.001). There were no differences in drawing completion times among the three motor states (p=0.509). In the OFF and ON states, patients drew slightly more impaired spirals in the afternoon whereas in the DYS state the spiral drawing performance was more impaired in the morning. Conclusion: It was found that when patients considered themselves as being dyskinetic spiral drawing was more impaired (nearly one unit change in a 0-10 scale) compared to when they considered themselves as being 'off' and 'on'. The spiral drawing at patients that self-assessed their motor state as dyskinetic was slightly more impaired in the morning hours, between 8 and 12 o'clock, a situation possibly caused by the morning dose effect.

The Optimal trigger speed of vehicle activated signs

The thesis aims to elaborate on the optimum trigger speed for Vehicle Activated Signs (VAS) and to study the effectiveness of VAS trigger speed on drivers’ behaviour. Vehicle activated signs (VAS) are speed warning signs that are activated by individual vehicle when the driver exceeds a speed threshold. The threshold, which triggers the VAS, is commonly based on a driver speed, and accordingly, is called a trigger speed. At present, the trigger speed activating the VAS is usually set to a constant value and does not consider the fact that an optimal trigger speed might exist. The optimal trigger speed significantly impacts driver behaviour. In order to be able to fulfil the aims of this thesis, systematic vehicle speed data were collected from field experiments that utilized Doppler radar. Further calibration methods for the radar used in the experiment have been developed and evaluated to provide accurate data for the experiment. The calibration method was bidirectional; consisting of data cleaning and data reconstruction. The data cleaning calibration had a superior performance than the calibration based on the reconstructed data. To study the effectiveness of trigger speed on driver behaviour, the collected data were analysed by both descriptive and inferential statistics. Both descriptive and inferential statistics showed that the change in trigger speed had an effect on vehicle mean speed and on vehicle standard deviation of the mean speed. When the trigger speed was set near the speed limit, the standard deviation was high. Therefore, the choice of trigger speed cannot be based solely on the speed limit at the proposed VAS location. The optimal trigger speeds for VAS were not considered in previous studies. As well, the relationship between the trigger value and its consequences under different conditions were not clearly stated. The finding from this thesis is that the optimal trigger speed should be primarily based on lowering the standard deviation rather than lowering the mean speed of vehicles. Furthermore, the optimal trigger speed should be set near the 85th percentile speed, with the goal of lowering the standard deviation.

Assessment of PD Speech Anomalies @ Home

Background: Voice processing in real-time is challenging. A drawback of previous work for Hypokinetic Dysarthria (HKD) recognition is the requirement of controlled settings in a laboratory environment. A personal digital assistant (PDA) has been developed for home assessment of PD patients. The PDA offers sound processing capabilities, which allow for developing a module for recognition and quantification HKD. Objective: To compose an algorithm for assessment of PD speech severity in the home environment based on a review synthesis. Methods: A two-tier review methodology is utilized. The first tier focuses on real-time problems in speech detection. In the second tier, acoustics features that are robust to medication changes in Levodopa-responsive patients are investigated for HKD recognition. Keywords such as Hypokinetic Dysarthria , and Speech recognition in real time were used in the search engines. IEEE explorer produced the most useful search hits as compared to Google Scholar, ELIN, EBRARY, PubMed and LIBRIS. Results: Vowel and consonant formants are the most relevant acoustic parameters to reflect PD medication changes. Since relevant speech segments (consonants and vowels) contains minority of speech energy, intelligibility can be improved by amplifying the voice signal using amplitude compression. Pause detection and peak to average power rate calculations for voice segmentation produce rich voice features in real time. Enhancements in voice segmentation can be done by inducing Zero-Crossing rate (ZCR). Consonants have high ZCR whereas vowels have low ZCR. Wavelet transform is found promising for voice analysis since it quantizes non-stationary voice signals over time-series using scale and translation parameters. In this way voice intelligibility in the waveforms can be analyzed in each time frame. Conclusions: This review evaluated HKD recognition algorithms to develop a tool for PD speech home-assessment using modern mobile technology. An algorithm that tackles realtime constraints in HKD recognition based on the review synthesis is proposed. We suggest that speech features may be further processed using wavelet transforms and used with a neural network for detection and quantification of speech anomalies related to PD. Based on this model, patients' speech can be automatically categorized according to UPDRS speech ratings.

Motion Cues Analysis for Parkinson Gait Recognition

Background: Previous assessment methods for PG recognition used sensor mechanisms for PG that may cause discomfort. In order to avoid stress of applying wearable sensors, computer vision (CV) based diagnostic systems for PG recognition have been proposed. Main constraints in these methods are the laboratory setup procedures: Novel colored dresses for the patients were specifically designed to segment the test body from a specific colored background. Objective: To develop an image processing tool for home-assessment of Parkinson Gait(PG) by analyzing motion cues extracted during the gait cycles. Methods: The system is based on the idea that a normal body attains equilibrium during the gait by aligning the body posture with the axis of gravity. Due to the rigidity in muscular tone, persons with PD fail to align their bodies with the axis of gravity. The leaned posture of PD patients appears to fall forward. Whereas a normal posture exhibits a constant erect posture throughout the gait. Patients with PD walk with shortened stride angle (less than 15 degrees on average) with high variability in the stride frequency. Whereas a normal gait exhibits a constant stride frequency with an average stride angle of 45 degrees. In order to analyze PG, levodopa-responsive patients and normal controls were videotaped with several gait cycles. First, the test body is segmented in each frame of the gait video based on the pixel contrast from the background to form a silhouette. Next, the center of gravity of this silhouette is calculated. This silhouette is further skeletonized from the video frames to extract the motion cues. Two motion cues were stride frequency based on the cyclic leg motion and the lean frequency based on the angle between the leaned torso tangent and the axis of gravity. The differences in the peaks in stride and lean frequencies between PG and normal gait are calculated using Cosine Similarity measurements. Results: High cosine dissimilarity was observed in the stride and lean frequencies between PG and normal gait. High variations are found in the stride intervals of PG whereas constant stride intervals are found in the normal gait. Conclusions: We propose an algorithm as a source to eliminate laboratory constraints and discomfort during PG analysis. Installing this tool in a home computer with a webcam allows assessment of gait in the home environment.

Short Communication: A compelling argument for the gravity p-median model

The p-median model is used to locate P facilities to serve a geographically distributed population. Conventionally, it is assumed that the population always travels to the nearest facility. Drezner and Drezner (2006, 2007) provide three arguments on why this assumption might be incorrect, and they introduce the extended the gravity p-median model to relax the assumption. We favour the gravity p-median model, but we note that in an applied setting, Drezner and Drezner’s arguments are incomplete. In this communication, we point at the existence of a fourth compelling argument for the gravity p-median model.

Methodological issues in applying Location Models to Rural areas

Location Models are usedfor planning the location of multiple service centers in order to serve a geographicallydistributed population. A cornerstone of such models is the measure of distancebetween the service center and a set of demand points, viz, the location of thepopulation (customers, pupils, patients and so on). Theoretical as well asempirical evidence support the current practice of using the Euclidian distancein metropolitan areas. In this paper, we argue and provide empirical evidencethat such a measure is misleading once the Location Models are applied to ruralareas with heterogeneous transport networks. This paper stems from the problemof finding an optimal allocation of a pre-specified number of hospitals in alarge Swedish region with a low population density. We conclude that the Euclidianand the network distances based on a homogenous network (equal travel costs inthe whole network) give approximately the same optimums. However networkdistances calculated from a heterogeneous network (different travel costs indifferent parts of the network) give widely different optimums when the numberof hospitals increases.  In terms ofaccessibility we find that the recent closure of hospitals and the in-optimallocation of the remaining ones has increased the average travel distance by 75%for the population. Finally, aggregation the population misplaces the hospitalsby on average 10 km.