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

Deposits of clastic carbonate-dominated (calciclastic) sedimentary slope systems in the rock record have been identified mostly as linearly-consistent carbonate apron deposits, even though most ancient clastic carbonate slope deposits fit the submarine fan systems better. Calciclastic submarine fans are consequently rarely described and are poorly understood. Subsequently, very little is known especially in mud-dominated calciclastic submarine fan systems. Presented in this study are a sedimentological core and petrographic characterisation of samples from eleven boreholes from the Lower Carboniferous of Bowland Basin (Northwest England) that reveals a >250 m thick calciturbidite complex deposited in a calciclastic submarine fan setting. Seven facies are recognised from core and thin section characterisation and are grouped into three carbonate turbidite sequences. They include: 1) Calciturbidites, comprising mostly of highto low-density, wavy-laminated bioclast-rich facies; 2) low-density densite mudstones which are characterised by planar laminated and unlaminated muddominated facies; and 3) Calcidebrites which are muddy or hyper-concentrated debrisflow deposits occurring as poorly-sorted, chaotic, mud-supported floatstones. These

Phd by thesis

Today's Web-enabled deluge of electronic data calls for automated methods of data analysis. Machine learning provides these, developing methods that can automatically detect patterns in data and then use the uncovered patterns to predict future data. This textbook offers a comprehensive and self-contained introduction to the field of machine learning, based on a unified, probabilistic approach. The coverage combines breadth and depth, offering necessary background material on such topics as probability, optimization, and linear algebra as well as discussion of recent developments in the field, including conditional random fields, L1 regularization, and deep learning. The book is written in an informal, accessible style, complete with pseudo-code for the most important algorithms. All topics are copiously illustrated with color images and worked examples drawn from such application domains as biology, text processing, computer vision, and robotics. Rather than providing a cookbook of different heuristic methods, the book stresses a principled model-based approach, often using the language of graphical models to specify models in a concise and intuitive way. Almost all the models described have been implemented in a MATLAB software package--PMTK (probabilistic modeling toolkit)--that is freely available online. The book is suitable for upper-level undergraduates with an introductory-level college math background and beginning graduate students.

Machine Learning : A Probabilistic Perspective

Data analysis plays an indispensable role for understanding various phenomena. Cluster analysis, primitive exploration with little or no prior knowledge, consists of research developed across a wide variety of communities. The diversity, on one hand, equips us with many tools. On the other hand, the profusion of options causes confusion. We survey clustering algorithms for data sets appearing in statistics, computer science, and machine learning, and illustrate their applications in some benchmark data sets, the traveling salesman problem, and bioinformatics, a new field attracting intensive efforts. Several tightly related topics, proximity measure, and cluster validation, are also discussed.

/pdf/survey-of-clustering-algorithms-6zezqhbjyo.pdf

Survey of clustering algorithms

We propose an heuristic algorithm that induces decision graphs from training sets using Rissanen's minimum description length principle to control the tradeoff between accuracy in the training set and complexity of the hypothesis description.

Inferring reduced ordered decision graphs of minimum description length

Although many algorithms for power estimation havebeen proposed to date, no comprehensive results havebeen presented on the actual complexity of power esti-mation problems.In this paper we select a number of relevant problemsand show that they belong to classes of high complexity.In particular we show that: a) for a given combinationalcircuit, the decision version of the peak power estimationproblem is NP-complete; b) average power estimation forcombinational circuits is #P-complete under the uniforminput distribution; c) the decision versionof the sequentialpower estimation problem is PSPACE-complete.We also address a numberof related problems and con-clude that even approximation algorithms for some ofthese problems are NP-hard. 1 Introduction Extensiveresearch has been done on algorithmsforpowerestimation of CMOS circuits. A number of problemslike peak power estimation and average power dissipa-tion in combinational and sequential circuits have beenaddressed using different methods. Although there ex-ists considerable agreement that efﬁcient (i.e., guaranteedpolynomial time) algorithms do not exist for the major-ity of these problems, no speciﬁc analysis of their actualcomplexity has been published.This works addresses this lack of speciﬁc results, andshows, using a relatively straightforward analysis, thatseveral power estimation problems belong to classes ofvery high complexity.The ﬁrst problem of interest we consideris the problemof peak power estimation. We show in section 3 that thedecision version of this problem is NP-complete.In sections 4 and 5, we consider the more general prob-lem of power estimation of combinational and sequentialcircuits.Existing algorithms for power estimation can be di-vided into two categories: simulation based methods andprobabilistic methods.Simulation based methods work by simulating a giveninput trace. The complexity of simulation based powerestimation methods is easily seen to be polynomial on thesize of the trace and on the size of the circuit. However,an accurate characterizationof the operatingconditionsofthe circuit require the existence of a trace that can be ex-ponentially large on the dimension of the circuit, therebymaking the complexity of these methodseffectively expo-nential on the size of the circuit. To avoid this exponentialblowup without incurring in a signiﬁcant loss of preci-sion, approximate methods like sequence compaction [8]and sequence synthesis [7] have been proposed.The alternative to this approach are probabilistic powerestimation methods. They do not explicitly require theexistence of a trace and, in principle, can be used to com-pute the power dissipation of circuits without the needfor an exponentially long description of the input stim-uli. These methods usually assume simpliﬁed models oftemporal and spatial correlations. One common simpli-ﬁcation is the assumption that the primary inputs are un-correlated in time and space. Both exact [5] and approxi-mate [2] algorithms, for this problem have been proposed.Methods that use a more accurate modeling of spatial andtemporal correlations have also been presented. One suchmethod [9] models the pairwise spatial correlations of theprimary inputs and propagates them through the circuit.A different approach that is able to take into account thefull set of temporal and spatial correlations at the inputshas also been proposed [3].Experimentally, it has been observed that all statisticalpower estimation methods become computationally veryexpensive as the size of the circuit grows. In sections 4and 5 we show that the problem of power estimation incombinational and sequential circuits is indeed inherentlycomplex, and that this complexity cannot be avoided bythe use of statistical power estimation methods. In fact,in section 6, we show, using a simple argument, that noefﬁcient approximation algorithms are likely to exist.

On The Complexity Of Power Estimation Problems

Haplotype inference from genotype data is a key step towards a better understanding of the role played by genetic variations on inherited diseases. One of the most promising approaches uses the pure parsimony criterion. This approach is called Haplotype Inference by Pure Parsimony (HIPP) and is NP-hard as it aims at minimising the number of haplotypes required to explain a given set of genotypes. The HIPP problem is often solved using constraint satisfaction techniques, for which the upper bound on the number of required haplotypes is a key issue. Another very well-known approach is Clark's method, which resolves genotypes by greedily selecting an explaining pair of haplotypes. In this work, we combine the basic idea of Clark's method with a more sophisticated method for the selection of explaining haplotypes, in order to explicitly introduce a bias towards parsimonious explanations. This new algorithm can be used either to obtain an approximated solution to the HIPP problem or to obtain an upper bound on the size of the pure parsimony solution. This upper bound can then used to efficiently encode the problem as a constraint satisfaction problem. The experimental evaluation, conducted using a large set of real and artificially generated examples, shows that the new method is much more effective than Clark's method at obtaining parsimonious solutions, while keeping the advantages of simplicity and speed of Clark's method.

Efficient and tight upper bounds for haplotype inference by pure parsimony using delayed haplotype selection

We study parallel and distributed compressed indexes. Compressed indexes are a new and functional way to index text strings. They exploit the compressibility of the text, so that their size is a function of the compressed text size. Moreover, they support a considerable amount of functions, more than many classical indexes. We make use of this extended functionality to obtain, in a shared-memory parallel machine, near-optimal speedups for solving several stringology problems. We also show how to distribute compressed indexes across several machines.

/pdf/parallel-and-distributed-compressed-indexes-24l7omgsxc.pdf

Parallel and distributed compressed indexes

Algorithms for the inference of association with sequential information have been proposed and used but are ineffective, in some cases, because too many candidate rules are extracted. Filtering the relevant ones is usually difficult and inefficient. In this work, we present an algorithm for the inference of temporal association rules that uses context-free grammars to restrict the search process, in order to filter, in an efficient and effective way, the associations discovered by the algorithm. Moreover, we present experimental results that empirically evaluate its performance using synthetic data.

Arlindo L. Oliveira

Papers

Inferring reduced ordered decision graphs of minimum description length

On The Complexity Of Power Estimation Problems

Efficient and tight upper bounds for haplotype inference by pure parsimony using delayed haplotype selection

Parallel and distributed compressed indexes

Using Context-Free Grammars to Constrain Apriori-based Algorithms for Mining Temporal Association Rules