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

MKT 6009 Marketing Internship (0 semester credit hours) Student gains experience and improves skills through appropriate developmental work assignments in a real business environment. Student must identify and submit specific business learning objectives at the beginning of the semester. The student must demonstrate exposure to the managerial perspective via involvement or observation. At semester end, student prepares an oral or poster presentation, or a written paper reflecting on the work experience. Student performance is evaluated by the work supervisor. Pass/Fail only. Prerequisites: (MAS 6102 or MBA major) and department consent required. (0-0) S MKT 6244 Digital Marketing Strategy (2 semester credit hours) Executive Education Course. The course explores three distinct areas within marketing and sales namely, digital marketing, traditional sales prospecting, and executive sales organization and strategy. The continuing convergence of the digital marketing and sales funnels has created a strategic continuum from digital lead generation to digital sales. The course identifies the current composition of this digital continuum while providing opportunities to evaluate sales and marketing digital strategies. Prerequisites: MKT 6301 and instructor consent required. (2-0) Y MKT 6301 (SYSM 6318) Marketing Management (3 semester credit hours) Overview of marketing management methods, principles and concepts including product, pricing, promotion and distribution decisions as well as segmentation, targeting and positioning. (3-0) S MKT 6309 Marketing Data Analysis and Research (3 semester credit hours) Methods employed in market research and data analysis to understand consumer behavior, customer journeys, and markets so as to enable better decision-making. Topics include understanding different sources of data, survey design, experiments, and sampling plans. The course will cover the techniques used for market sizing estimation and forecasting. In addition, the course will cover the foundational concepts and techniques used in data visualization and \"story-telling\" for clients and management. Corequisites: MKT 6301 and OPRE 6301. (3-0) Y MKT 6310 Consumer Behavior (3 semester credit hours) An exposition of the theoretical perspectives of consumer behavior along with practical marketing implication. Study of psychological, sociological and behavioral findings and frameworks with reference to consumer decision-making. Topics will include the consumer decision-making model, individual determinants of consumer behavior and environmental influences on consumer behavior and their impact on marketing. Prerequisite: MKT 6301. (3-0) Y MKT 6321 Interactive and Digital Marketing (3 semester credit hours) Introduction to the theory and practice of interactive and digital marketing. Topics covered include: online-market research, consumer behavior, conversion metrics, and segmentation considerations; ecommerce, search and display advertising, audiences, search engine marketing, email, mobile, video, social networks, and the Internet of Things. (3-0) T MKT 6322 Internet Business Models (3 semester credit hours) Topics to be covered are: consumer behavior on the Internet, advertising on the Internet, competitive strategies, market research using the Internet, brand management, managing distribution and supply chains, pricing strategies, electronic payment systems, and developing virtual organizations. Further, students learn auction theory, web content design, and clickstream analysis. Prerequisite: MKT 6301. (3-0) Y MKT 6323 Database Marketing (3 semester credit hours) Techniques to analyze, interpret, and utilize marketing databases of customers to identify a firm's best customers, understanding their needs, and targeting communications and promotions to retain such customers. Topics

Marketing for management.

Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI

Genetic programming (GP) is a systematic, domain-independent method for getting computers to solve problems automatically starting from a high-level statement of what needs to be done. Using ideas from natural evolution, GP starts from an ooze of random computer programs, and progressively refines them through processes of mutation and sexual recombination, until high-fitness solutions emerge. All this without the user having to know or specify the form or structure of solutions in advance. GP has generated a plethora of human-competitive results and applications, including novel scientific discoveries and patentable inventions. This unique overview of this exciting technique is written by three of the most active scientists in GP. See www.gp-field-guide.org.uk for more information on the book.

/pdf/a-field-guide-to-genetic-programming-3wkkmsw81n.pdf

A Field Guide to Genetic Programming

This paper introduces a novel focus measure for Shape From Focus (SFF) based on 1-dimensional Discrete Cosine Transform (DCT) Instead of locally computing the focus quality in a window, we constructed a vector consisting of seven pixels for each pixel in the image volume The new neighborhood incorporates the effect on focus value of pixels from 3-dimensions In transform domain, to exploit the variation in energy, each AC coefficient is subtracted from the DC component The focus value is then computed considering the energy of the modified AC components The proposed method is simple but effective Experimental results demonstrate the effectiveness of the new focus measure by using two different image sequences

A Transformed Domain based Novel Focus Measure for 3D Shape Recovery

Diabetic retinopathy (DR) is a diabetes complication that can cause vision loss among patients due to damage to blood vessels in the retina. Early retinal screening can avoid the severe consequences of DR and enable timely treatment. Nowadays, researchers are trying to develop automated deep learning-based DR segmentation tools using retinal fundus images to help Ophthalmologists with DR screening and early diagnosis. However, recent studies are unable to design accurate models due to the unavailability of larger training data with consistent and fine-grained annotations. To address this problem, we propose a semi-supervised multitask learning approach that exploits widely available unlabelled data (i.e., Kaggle-EyePACS) to improve DR segmentation performance. The proposed model consists of novel multi-decoder architecture and involves both unsupervised and supervised learning phases. The model is trained for the unsupervised auxiliary task to effectively learn from additional unlabelled data and improve the performance of the primary task of DR segmentation. The proposed technique is rigorously evaluated on two publicly available datasets (i.e., FGADR and IDRiD) and results show that the proposed technique not only outperforms existing state-of-the-art techniques but also exhibits improved generalisation and robustness for cross-data evaluation. 

/pdf/ssmd-unet-semi-supervised-multi-task-decoders-network-for-khg5su1p.pdf

SSMD-UNet: semi-supervised multi-task decoders network for diabetic retinopathy segmentation

Prediction based on the hydrophobicity of the protein yields potentially good classification rate as compared to the other compositions for G-Proteins coupled receptor (GPCR's) families and their respective subfamilies. In the current study, we make use of the hydrophobicity of the proteins in order to obtain a fourier spectrum of the protein sequence, which is then used for classification purpose. The classification of 17 GPCR subfamilies is based on Nearest Neighbor (NN) method, which is employed at two levels. At level-1 classification, the GPCR super-family is recognized and at level-2, the respective sub-families for the predicted super-family are classified. As against Support Vector Machine (SVM), NN approach has shown better performance using both jackknife and independent data set testing. The results are formulated using three performance measures, the Mathew's Correlation Coefficient (MCC), overall accuracy (ACC) and reliability (R) on both training and independent data sets. Comparison of our results is carried out with the overall class accuracies obtained for super-families using existing technique. The multilevel classifier has shown promising performance and has achieved overall ACC and MCC of 97.02% and 0.95 using jackknife test, and 87.50 % and 0.85 for independent data set test respectively.

Using multi level nearest neighbor classifiers for G-protein coupled receptor sub-families prediction

Human immune system plays a vital role in cancer prevention. Tumor Infiltrating Lymphocytes (TILs) are specialized immune cells that infiltrate tumor region. Studies have shown that the presence of TILs reflect towards cancer prognosis. The amount of TILs in tumor stroma can be counted manually under a microscope by an expert pathologist. This manual task is very laborious and time-consuming, and may face human subjectivity. Therefore, an automated system is required to effectively detect lymphocytes and to overcome the burden on pathologist. Automation of lymphocyte detection is a challenging task, due to its morphological variations, clustered appearance and presence of artifacts. To address these challenges, we have designed a computer aided diagnostic system called Pyramid Vision Transformer Channel Boosted Lymphocyte Detection (PVTCB-Lymph-Det) by exploiting deep CNN architectures based on channel boosting. In this work, we have incorporated the idea of channel boosting to enhance the learning capacity of the proposed model. The proposed network is based on a transformer and Convolutional Block Attention Mechanism (CBAM) with ResNet-50 to perform feature extraction. The proposed technique with Channel Boosting has performed well for lymphocyte detection and has an F-score of 88.92%. Moreover, PVTCB-Lymph-Det can be adapted to detect lymphocytes and may aid pathologists in solving medical diagnostic problems.

Channel boosting based detection and segmentation for cancer analysis in histopathological images.

Vision transformers have recently become popular as a possible alternative to convolutional neural networks (CNNs) for a variety of computer vision applications. These vision transformers due to their ability to focus on global relationships in images have large capacity, but may result in poor generalization as compared to CNNs. Very recently, the hybridization of convolution and self-attention mechanisms in vision transformers is gaining popularity due to their ability of exploiting both local and global image representations. These CNN-Transformer architectures also known as hybrid vision transformers have shown remarkable results for vision applications. Recently, due to the rapidly growing number of these hybrid vision transformers, there is a need for a taxonomy and explanation of these architectures. This survey presents a taxonomy of the recent vision transformer architectures, and more specifically that of the hybrid vision transformers. Additionally, the key features of each architecture such as the attention mechanisms, positional embeddings, multi-scale processing, and convolution are also discussed. This survey highlights the potential of hybrid vision transformers to achieve outstanding performance on a variety of computer vision tasks. Moreover, it also points towards the future directions of this rapidly evolving field.

Asifullah Khan

Papers

A Transformed Domain based Novel Focus Measure for 3D Shape Recovery

SSMD-UNet: semi-supervised multi-task decoders network for diabetic retinopathy segmentation

Using multi level nearest neighbor classifiers for G-protein coupled receptor sub-families prediction

Channel boosting based detection and segmentation for cancer analysis in histopathological images.

A survey of the Vision Transformers and its CNN-Transformer based Variants