Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Machine learning

The Analysis of Time Series: An Introduction, 4th edn. By C. Chatfield. ISBN 0 412 31820 2. Chapman and Hall, London, 1989. 242 pp. £13.50.

The Analysis of Time Series: An Introduction

Recently, many efforts have been made to develop more efficient Inter-Vehicle Communication (IVC) protocols for on-demand route planning according to observed traffic congestion or incidents, as well as for safety applications. Because practical experiments are often not feasible, simulation of network protocol behavior in Vehicular Ad Hoc Network (VANET) scenarios is strongly demanded for evaluating the applicability of developed network protocols. In this work, we discuss the need for bidirectional coupling of network simulation and road traffic microsimulation for evaluating IVC protocols. As the selection of a mobility model influences the outcome of simulations to a great extent, the use of a representative model is necessary for producing meaningful evaluation results. Based on these observations, we developed the hybrid simulation framework Veins (Vehicles in Network Simulation), composed of the network simulator OMNeT++ and the road traffic simulator SUMO. In a proof-of-concept study, we demonstrate its advantages and the need for bidirectionally coupled simulation based on the evaluation of two protocols for incident warning over VANETs. With our developed methodology, we can advance the state-of-the-art in performance evaluation of IVC and provide means to evaluate developed protocols more accurately.

/pdf/bidirectionally-coupled-network-and-road-traffic-simulation-59x3tw1e3m.pdf

Bidirectionally Coupled Network and Road Traffic Simulation for Improved IVC Analysis

Vehicular Ad-hoc Networks (VANETs) have been recently attracting an increasing attention from both research and industry communities. One of the challenges posed by the study of VANETs is the definition of a vehicular mobility model providing an accurate and realistic vehicular mobility description at both macroscopic and microscopic levels. Another challenge is to be able to dynamically alter this vehicular mobility as a consequence of the vehicular communication protocols. Many mobility models have been developed by the community in order to solve these two issues. However, due to the large number of available models claiming to be adapted to vehicular traffic, and also due to their different and somehow incomparable features, understanding their true characteristics, their degree of realism with respect to vehicular mobility, and real capabilities is a hard task. In this survey, we first introduce a framework that proposes a guideline for the generation of vehicular mobility models. Then, we illustrate the different approaches chosen by the community for the development of vehicular mobility models and their interactions with network simulators. Finally, we propose an overview and taxonomy of a large range of mobility models available for vehicular ad hoc networks. The objective is to provide readers with a guideline to easily understand and objectively compare the different models, and eventually identify the one required for their needs.

/pdf/mobility-models-for-vehicular-ad-hoc-networks-a-survey-and-1g8mz9v0dy.pdf

Mobility models for vehicular ad hoc networks: a survey and taxonomy

Inter-vehicle communication (IVC) systems (i.e., systems not relying on roadside infrastructure) have the potential to radically improve the safety, efficiency, and comfort of everyday road travel. Their main advantage is that they bypass the need for expensive infrastructure; their major drawback is the comparatively complex networking protocols and the need for significant penetration before their applications can become effective. In this article we present several major classes of applications and the types of services they require from an underlying network. We then proceed to analyze existing networking protocols in a bottom-up fashion, from the physical to the transport layers, as well as security aspects related to IVC systems. We conclude the article by presenting several projects related to IVC as well as a review of common performance evaluation techniques for IVC systems.

/pdf/inter-vehicle-communication-systems-a-survey-4dcec4ylxn.pdf

Inter-vehicle communication systems: a survey

There are growing concerns on traffic congestion, climate change and parking problems in major cities. Faced with these concerns, policy makers have sought sustainable transportation options including electric vehicle sharing programs (EVSPs). The city of Seoul with 10 million people also has recently launched an EVSP to provide citizens with an alternative travel mode. This study attempts to explore factors affecting the EVSP participants’ attitudes about car ownership and program participation. To do this, a web-based survey was conducted for the participants of the Seoul EVSP, asking their satisfaction levels for the components of the EVSP. Then, using 533 responses of 1772 EVSP members (a response rate of 30%), ordered probit models were developed for three types of attitudes: (1) willingness to dispose of a car, (2) willingness to purchase an EV and (3) willingness to continue participating in the EVSP. The estimated models suggested that participants’ social and economic perspectives were the most important factors affecting the participants’ attitudes. In addition, the attitudes varied depending on personal characteristics such as gender, age and income. Although this study was conducted in the early stage of an EVSP, its results are expected to provide insights into a better EVSP design.

Factors affecting electric vehicle sharing program participants’ attitudes about car ownership and program participation

Exploitation of in-vehicle information technology (e.g., mobile computing and wireless communications) in surface transportation systems is a clearly emerging trend. Equipping vehicles with computing, communication, and sensing capabilities presents significant opportunities for a vast array of transportation services. Vehicle-to-vehicle (V2V) communication may be considered for applications such as incident detection, crash reporting, traveler information dissemination, and network operations. In-vehicle computing systems facilitate the customization of information services to the needs and characteristics of individual travelers. In addition, these systems allow coverage to extend beyond areas where roadside equipment has been placed. This study provides the initial investigation needed to test the feasibility of these advanced communication networks. Several observations may be drawn from the study. First, V2V communication is a feasible way to propagate information along the I-75 freeway in the Atlant...

Efficiency of Simulated Vehicle-to-Vehicle Message Propagation in Atlanta, Georgia, I-75 Corridor

This study identifies the largest producers of CO2 emissions in the Seoul metropolis area, focusing on land-based passenger transportation activities as revealed by the 2006 household travel diary survey. Combining modal CO2 intensities and travel distances, the amount of daily individual CO2 emissions is estimated. The cumulative distribution of the CO2 emissions revealed that the top 10% emitters in Seoul, who produce daily about 4.5 kg of CO2, were responsible for 63% of transportation’s emissions. CO2 emissions are found to be strongly associated with emitters’ socioeconomic characteristics, with adult males being the largest emitters producing about 3.9 kg of CO2 daily.

Who produces the most CO2 emissions for trips in the Seoul metropolis area

Exploring factors associated with commute mode choice: An application of city-level general social survey data

This paper presents a detailed measurement study of short range communications between vehicles and between vehicles and roadside stations in a realistic highway scenario. We show the expected wireless communication characteristics in a driving environment and identify factors that significantly affect communication performance. We also illustrate the benefits of multi-hop communication in improving communication performance.

Joonho Ko

Papers

Factors affecting electric vehicle sharing program participants’ attitudes about car ownership and program participation

Efficiency of Simulated Vehicle-to-Vehicle Message Propagation in Atlanta, Georgia, I-75 Corridor

Who produces the most CO2 emissions for trips in the Seoul metropolis area

Exploring factors associated with commute mode choice: An application of city-level general social survey data

An empirical study of short range communications for vehicles