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

Data Mining Practical Machine Learning Tools and Techniques

The modern science of networks has brought significant advances to our understanding of complex systems. One of the most relevant features of graphs representing real systems is community structure, or clustering, i. e. the organization of vertices in clusters, with many edges joining vertices of the same cluster and comparatively few edges joining vertices of different clusters. Such clusters, or communities, can be considered as fairly independent compartments of a graph, playing a similar role like, e. g., the tissues or the organs in the human body. Detecting communities is of great importance in sociology, biology and computer science, disciplines where systems are often represented as graphs. This problem is very hard and not yet satisfactorily solved, despite the huge effort of a large interdisciplinary community of scientists working on it over the past few years. We will attempt a thorough exposition of the topic, from the definition of the main elements of the problem, to the presentation of most methods developed, with a special focus on techniques designed by statistical physicists, from the discussion of crucial issues like the significance of clustering and how methods should be tested and compared against each other, to the description of applications to real networks.

/pdf/community-detection-in-graphs-34rogm3r6p.pdf

Community detection in graphs

Statistics for Spatial Data.

It is crucial for cellular data network operators to understand the service quality perceived by its customers. The state-of-art systems deployed in cellular networks mostly report service quality aggregated on cell site level, which is typically an aggregation of tens or hundreds of customers depending on the locations of the cell sites. In this paper, we propose to enhance the measurement of customer-perceived service quality by leveraging M2M devices as sensors in the field, which provide an unprecedented opportunity for cellular network operators to measure what end-users experience with better accuracy and coverage. Our approach is to identify a set of M2M devices which are stationary and communicate continuously over the cellular network over an indefinite period of time. We use these M2M devices to estimate the customer-perceived service quality during cell site outages. We implement our methodology as a system called M2MScan and evaluate M2MScan with both synthetic outages and real outages from a large-scale operational cellular network. To the best of our knowledge, this is the first work that employs M2M devices to measure the service quality perceived by customers in operational cellular networks at a large scale.

Monitoring quality-of-experience for operational cellular networks using machine-to-machine traffic

Streaming video applications require high bandwidth for desired quality of experience (QoE), and they are driving rapid growth of mobile data traffic in cellular networks. Currently, cellular networks provide best-effort services to most user data applications. When there is congestion at the base station, streaming video applications will experience degraded QoE. In this paper, we take a cellular service provider’s perspective and propose a premium service for improving QoE of streaming video applications. We design and implement a network adaptation scheme called SHADE, which allocates limited transmission resources at the base station among applications smartly, by (i) selecting a candidate bitrate for each streaming video application, and (ii) maintaining the downlink throughput at this targeted bitrate for better QoE, while still using the Non-Guaranteed Bit Rate traffic class which is suitable for high bit-rate streaming video. We demonstrate that SHADE can achieve this with high network utilization and improve QoE for streaming video applications, and with bounded negative performance impact to other applications. Our extensive experiments show that SHADE can significantly improve three key streaming video application QoE metrics simultaneously (up to 10 times improvement), compared to current practice. We discuss how cellular carriers and equipment vendors can adopt SHADE without major changes to current cellular network implementation.

Enabling Premium Service for Streaming Video in Cellular Networks

This letter presents a dual-band lowpass–bandpass filter utilizing dual-mode unit cells. The first mode can be tuned without affecting the second mode. The I/O coupled lines for the bandpass response introduce a transmission zero (TZ) to improve the selectivity of the lowpass response. Indeed, the developed improved unit cell provides a frequency ratio ranging from 1 to 4 and a fractional bandwidth of up to 60%. Additionally, a seven-section lowpass–bandpass filter is designed and fabricated, affording a flexible passband allocation, where the lowpass response’s cutoff frequency and roll-off rate are 0.917 GHz and 1055 dB/GHz, respectively. The center frequency and fractional bandwidth of the bandpass response are 1.718 GHz and 13%, respectively, and the filter measures only $0.139\lambda _{g} \times 0.046\lambda _{g}$ .

High-Selectivity Lowpass–Bandpass Filters With Flexible Passband Allocation

—The Internet carries a vast amount and a wide rangeof content. Some of this content is more popular, and accessedmore frequently, than others. The popularity of content couldbe quite ephemeral - e.g., a Web ﬂash crowd - or much morepermanent - e.g., google.com’s banner. A direct consequence ofthe skew in popularity is that, at any time, a fraction of theinformation carried over the Internet is redundant.We make two contributions in this paper. First, we study thefundamental properties of the redundancy in the informationcarried over the Internet, with a focus on network edges. Wecollect trafﬁc traces at two network edge locations – a largeuniversity’s access link serving roughly 50,000 users, and a tier-1ISP network link connected to a large data center. We conductseveral analyses over this data: What fraction of bytes areredundant? What is the frequency at which strings of bytes re peatacross different packets? What is the overlap in the informa tionaccessed by distinct groups of end-users?Second, we leverage our measurement observations in thedesign of a family mechanisms for eliminating redundancy innetwork trafﬁc and improving the overall network performan ce.The mechanisms we proposed can improve the available capacityof single network links as well as balance load across multiplenetwork links.

Computer Sciences Department Understanding and Exploiting Network Traffic Redundancy

Multiple harmonics are suppressed by reduced external couplings and a symmetrical layout. The reduced external couplings are implemented by employing dual-hairpin resonators that simultaneously cancel the electric and magnetic fields. The symmetrical layout is utilized to additionally eliminate the spurious modes of substructure resonators. For validation, an eight-order superconducting filter incorporating two groups of stagger-tuned dual-hairpin resonators is designed and fabricated. The filter demonstrates excellent harmonic rejection suppressing six harmonics, extending the stopband to 17.4<italic>f</italic>0 with a rejection of 60 dB, highlighting straightforward synthesis and compact size (0.206<italic>λg</italic> × 0.046<italic>λg</italic>).

Jia Wang

Papers

Monitoring quality-of-experience for operational cellular networks using machine-to-machine traffic

Enabling Premium Service for Streaming Video in Cellular Networks

High-Selectivity Lowpass–Bandpass Filters With Flexible Passband Allocation

Computer Sciences Department Understanding and Exploiting Network Traffic Redundancy

Ultrawide Stopband Superconducting Filters With Multiple Harmonics Suppression