There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality.

Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

I and i

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

Wireless indoor positioning systems have become very popular in recent years. These systems have been successfully used in many applications such as asset tracking and inventory management. This paper provides an overview of the existing wireless indoor positioning solutions and attempts to classify different techniques and systems. Three typical location estimation schemes of triangulation, scene analysis, and proximity are analyzed. We also discuss location fingerprinting in detail since it is used in most current system or solutions. We then examine a set of properties by which location systems are evaluated, and apply this evaluation method to survey a number of existing systems. Comprehensive performance comparisons including accuracy, precision, complexity, scalability, robustness, and cost are presented.

Survey of Wireless Indoor Positioning Techniques and Systems

The PASCAL Visual Object Classes Challenge

An overview of the self-organizing map algorithm, on which the papers in this issue are based, is presented in this article.

The Self-Organizing Map

The diversity of applications, scarce resources, and large scale set demanding requirements for Wireless Sensor Networks (WSN). All requirements cannot be fulfilled by a general purpose WSN, but a development of application specific WSNs is needed. We present a novel WIreless SEnsor NEtwork Simulator (WISENES) framework for rapid design, simulation, evaluation, and implementation of both single nodes and large WSNs. New WSN design starts from high level Specification and Description Language (SDL) model, which is simulated and implemented on a prototype through code generation. One of the novel features is the back-annotation of measured values from physical prototypes to SDL model. The scalability and performance of WISENES have been evaluated with TUTWSN that is a very energy efficient new WSN. The results show only 6.7 percent difference between modeled and measured TUTWSN prototype energy consumption. Thus, WISENES hastens the development of WSN protocols and their evaluation in large networks.

High abstraction level design and implementation framework for wireless sensor networks

Resource constrained Wireless Sensor Networks WSNs require an automated firmware updating protocol for adding new features or error fixes. Reprogramming nodes manually is often impractical or even impossible. Current update protocols require a large external memory or external WSN transport protocol. This paper presents the design, implementation, and experiments of a Program Image Dissemination Protocol PIDP for autonomous WSNs. It is reliable, lightweight and it supports multi-hopping. PIDP does not require external memory, is independent of the WSN implementation, transfers firmware, and reprograms the whole program image. It was implemented on a node platform with an 8-bit microcontroller and a 2.4 GHz radio. Implementation requires 22 bytes of data memory and less than 7 kilobytes of program memory. PIDP updates 178 nodes within 5 hours. One update consumes under 1â° of the energy of two AA batteries.

Design and Implementation of a Firmware Update Protocol for Resource Constrained Wireless Sensor Networks

This paper introduces a model that can be used to share link capacity among customers under different kind of traffic conditions. This model is suitable for different kind of networks like the 4G networks (fast wireless access to wired network) to support connections of given duration that requires a certain quality of service. We study different types of network traffic mixed in a same communication link. A single link is considered as a bottleneck and the goal is to find customer traffic profiles that maximizes the revenue of the link. Presented allocation system accepts every calls and there is not absolute blocking, but the offered data rate/user depends on the network load. Data arrival rate depends on the current link utilization, user's payment (selected CoS class) and delay. The arrival rate is (i) increasing with respect to the offered data rate, (ii) decreasing with respect to the price, (iii) decreasing with respect to the network load, and (iv) decreasing with respect to the delay. As an example, explicit formula obeying these conditions is given and analyzed.

Link allocation and revenue optimization for future networks

Evolving video coding standards demand functional flexibility for implementations, not only at design time but also after fabrication. This paper presents a System-on-Chip design approach with a feasible combination of performance, scalability, programmability, area efficiency, and design time effort for a video encoder. The encoder is based on a homogeneous master-slave processor architecture. Each slave encodes a part of the frame in the Single Program Multiple Data (SPMD) data parallel model. Both shared and distributed memory architectures are presented. Design effort is reduced by identical program codes, automated assembly of software and hardware modules independent of the number and type of processors, as well as our flexible on-chip communication network called Heterogeneous IP Block Interconnection (HIBI). A case study implementation with two to ten simple ARM7 processors, 32-bit HIBI bus and non-optimized processor-independent software gives the performance from 6 to 53 fps for QCIF. The whole encoder area ranges from 173 to 770 kgates excluding the memories. The relation scales reasonably well to systems with more powerful processors and optimized code. The optimization of the communication network shows that with more than six slaves even a serial HIBI connection with 100 MHz speed is feasible. HIBI and the parallelization approach allow exploration and optimization of the communication both at the application and architecture layers.

Scalable Architecture for SoC Video Encoders

A Multiprocessor System-on-Chip (MPSoC) may contain hundreds of processing elements (PEs) and thousands of tasks but design productivity is lagging the evolution of HW platforms. One problem is application task mapping, which tries to find a placement of tasks onto PEs which optimizes several criteria such as application runtime, intertask communication, memory usage, energy consumption, real-time constraints, as well as area in case that PE selection or buffer sizing are combined with the mapping procedure. Among optimization algorithms for the task mapping, we focus in this paper on Simulated Annealing (SA) heuristics. We present a literature survey and 5 general recommendations for reporting heuristics that should allow disciplined comparisons and reproduction by other researchers. Most importantly, we present our findings about SA parameter selection and 7 guidelines for obtaining a good trade-off made between solution quality and algorithm's execution time. Notably, SA is compared against global optimum. Thorough experiments were performed with 2---8 PEs, 11---32 tasks, 10 graphs per system, and 1000 independent runs, totaling over 500 CPU days of computation. Results show that SA offers 4---6 orders of magnitude reduction is optimization time compared to brute force while achieving high quality solutions. In fact, the globally optimum solution was achieved with a 1.6--90 % probability when problem size is around 1e9---4e9 possibilities. There is approx. 90 % probability for finding a solution that is at most 18 % worse than optimum.

/pdf/recommendations-for-using-simulated-annealing-in-task-1d9vma31cx.pdf

Timo Hämäläinen

Papers

High abstraction level design and implementation framework for wireless sensor networks

Design and Implementation of a Firmware Update Protocol for Resource Constrained Wireless Sensor Networks

Link allocation and revenue optimization for future networks

Scalable Architecture for SoC Video Encoders

Recommendations for using Simulated Annealing in task mapping