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

Co-Planar Stereotaxic Atlas of the Human Brain—3-Dimensional Proportional System: An Approach to Cerebral Imaging, J. Talairach, P. Tournoux. Georg Thieme Verlag, New York (1988), 122 pp., 130 figs. DM 268

Learning and then recognizing a route, whether travelled during the day or at night, in clear or inclement weather, and in summer or winter is a challenging task for state of the art algorithms in computer vision and robotics. In this paper, we present a new approach to visual navigation under changing conditions dubbed SeqSLAM. Instead of calculating the single location most likely given a current image, our approach calculates the best candidate matching location within every local navigation sequence. Localization is then achieved by recognizing coherent sequences of these “local best matches”. This approach removes the need for global matching performance by the vision front-end - instead it must only pick the best match within any short sequence of images. The approach is applicable over environment changes that render traditional feature-based techniques ineffective. Using two car-mounted camera datasets we demonstrate the effectiveness of the algorithm and compare it to one of the most successful feature-based SLAM algorithms, FAB-MAP. The perceptual change in the datasets is extreme; repeated traverses through environments during the day and then in the middle of the night, at times separated by months or years and in opposite seasons, and in clear weather and extremely heavy rain. While the feature-based method fails, the sequence-based algorithm is able to match trajectory segments at 100% precision with recall rates of up to 60%.

SeqSLAM : visual route-based navigation for sunny summer days and stormy winter nights

Mobile-health

The relation between arterial function indices, such as pulse wave velocity and augmentation index with parameters derived from input impedance analysis, is still incompletely understood. Carotid pressure, central flow waveforms, and pulse wave velocity were noninvasively acquired in 2026 apparently healthy, middle-aged subjects (1052 women and 974 men) 35 to 55 years old at inclusion. Input and characteristic impedance, reflection coefficient, the ratio of backward-to-forward pressure amplitude (reflection magnitude), and augmentation index were derived. Pulse wave velocity increased by 15% (from 6.1 to 7.0 m/s) both in men and women. In qualitative terms, input impedance evolved from a pattern indicative of wave transmission and reflection to a pattern more compatible with a windkessel-like system. In women, a decrease in total arterial compliance led to an increased input impedance in the low frequency range, whereas few changes were observed in men. Characteristic impedance did not change with age in women and even decreased in men (P<0.001) and could not be identified as the primary determinant of central pulse pressure. Augmentation index increased with age, as was expected, and was systematically higher in women (P<0.001). Reflection coefficient and reflection magnitude increased with age (P<0.001) without gender differences. We conclude that, in healthy middle-aged subjects, the age-related increase in arterial stiffness (pulse wave velocity) is not fully paralleled by an increase in arterial impedance, suggesting a role for age-dependent modulation of aortic cross-sectional area. Wave reflection increases with age and is not higher in women than in men.

Noninvasive (input) impedance, pulse wave velocity, and wave reflection in healthy middle-aged men and women. Editorial Commentary

This paper presents an electronic travel aid for blind people to navigate safely and quickly, an obstacle detection system using ultrasonic sensors and USB camera based visual navigation has been considered. The proposed system detects the obstacles up to 300 cm via sonar and sends feedback (beep sound) to inform the person about its location. In addition to this, an USB webcam is connected with eBox 2300™ Embedded System for capturing the field of view of the user, which is used for finding the properties of the obstacle in particular, in the context of this work, locating a human being. Identification of human presence is based on face detection and cloth texture analysis. The major constraints for these algorithms to run on Embedded System are small image frame (160×120) having reduced faces, limited memory and very less processing time available to achieve real time image processing requirements. The algorithms are implemented in C++ using Visual Studio 5.0 IDE, which runs on Windows CE™ environment.

An electronic travel aid for navigation of visually impaired persons

Pose Depth Volume extraction from RGB-D streams for frontal gait recognition

Way-finding Electronic Bracelet (WEB) for visually impaired (subject) is a portable embedded system for obstacle detection and way-finding. This MSP430G2553 processor based real-time system employs single Maxbotix MB1340 ultrasonic transceiver mounted on customized circular bracelet for detecting obstacle in the range 20 to 600 centimetres. Half duplex wireless communication is used for invoking vibrotactile and audio cues at receiver side. Maintaining a safety margin distance in front, left and right direction, WEB dynamically calculates obstacle distance (if any). Using on-demand hand movements, subject can understand surrounding situations and can perform successful way-finding. WEB system is available with choice of optimum hardware as per wearable comfort and requirement of subject. Preliminary trials on blindfolded subjects, WEB demonstrated substantial potential for cost-effective wearable real-time system with minimum physical interface for mobility of visually impaired people.

Way-finding Electronic Bracelet for visually impaired people

Non-invasive and cuff-less measurement of arterial blood pressure (BP) is desirable for continuous patient monitoring. Among the various possible techniques pulse transit time (PTT) based approach for estimation of BP is the most promising one. But change in BP is reflected both in PTT and pre-ejection period (PEP). We propose to measure PTT using multiple PPG and PPG-ECG combination recorded at two different sites along the same artery and argue that this approach will eliminate a number of assumptions like constant PEP, involved in the most common method of measuring PTT using QRS complex as timing reference. This approach can help in finding an appropriate BP calibration methodology since true PTT can be measured by this technique.

Cuff-Less Estimation of Blood Pressure Using Pulse Transit Time and Pre-ejection Period

Security of telemedicine applications is not often given adequate importance by the developers and healthcare administrators primarily to reduce cost. Though some security safeguards are employed by these applications to comply with existing medical data security and privacy regulations, these are not adequate in todaypsilas context. Moreover, in a Web-based application environment not only the data but also the application itself is vulnerable to attackers. Keeping these concerns in mind, we present the design of a Web-based, four-tier telemedicine system named iMedik which is accessible over desktops as well as handheld devices. We have illustrated how the proposed system differs from existing three-tier Web applications. The compliance status of the application with HIPAA Security Guidelines has also been noted. The security measures described in our approach look into the four-tier architecture from an attackerpsilas viewpoint and present a simple road map for developing secure e-health application with anywhere, anytime availability.

Jayanta Mukhopadhyay

Papers

An electronic travel aid for navigation of visually impaired persons

Pose Depth Volume extraction from RGB-D streams for frontal gait recognition

Way-finding Electronic Bracelet for visually impaired people

Cuff-Less Estimation of Blood Pressure Using Pulse Transit Time and Pre-ejection Period

Security analysis and implementation of web-based telemedicine services with a four-tier architecture