Blog by Akash Mavle

  Classical Statistics in University Under-graduate courses or even Graduate courses starts with descriptive statistics and then moves into distribution fitting and then all the way to complex multivariate analysis. Essentially covering hypothesis testing, correlation, regression , factor analysis and Principal Component analysis. Statistics assumes a lot of a-priori knowledge about the data and its properties and does not necessarily cover a lot of trial and error or even tinkering. Machine Learning in new age loo k s at wide array of techniques and algorithms which themselves learn from the data. Deep Machine Learning, Supervised Learning and Reinforcement Learning covers very interesting algorithm which learn themselves from the wide array of data. So data becomes input and model becomes output. This happens without any human intervention ( except in supervised learning). This is the real beauty of ML over conventional statistics. Although new age ML ( covering C...

  What machine learning algorithm should I use? •The answer to the question “What machine learning algorithm should I use?” is always “It depends.” •It depends on the size, quality, and nature of the data. •It depends on what you want to do with the answer. •It depends on how the math of the algorithm was translated into instructions for the computer you are using. And it depends on how much time you have. Even the most experienced data scientists can’t tell which algorithm will perform best before trying them Flavors of machine learning Supervised •Supervised learning a l gorithms make predictions based on a set of examples. For instance, historical stock prices can be used to hazard guesses at future prices. Each example used for training is labeled with the value of interest — in this case the stock price. •A supervised learning algorithm looks for patterns in those value labels. It can use any information that might be relevant — the day of the week, the season, the c...

  Machine Learning has gone through multiple waves of its adoption. Over the years data availability has increased exponentially. At the same time computer power has multiplied according to Moore’s law creating multitude of opportunities for machine learning. Machine Learning is going through rapid evolution from basic ma c hine learning ( hard wired techniques which came initially from statistics) to advanced machine learning to deep machine learning and the whole umbrella of algorithms in Artificial Intelligence covering not only learning, but problem representations, complex state-space models, reasoning, perception, thoughts, emotions and all the way up-to theorem proving, problem solving, search to advanced tasks like planning. Every step in right direction has made overall machine learning algorithms as one of the most talked phenomenon only with ML Scientists but practitioners as well as lay people. I would like to elaborate on some of the myths about Machine Le...

  Instrumentation and Control covers broad range of engineering techniques, practices covering Feedback Control System, Individual Controls loops, Programmable Logic Controllers (PLCs), Supervisory Control and Data Acquisition(SCADA) and specialised large scale control systems known as DCS ( Distributed Control Systems). Sensor and Transducers along with Actuators are other components in Instrumentation. Control Theory covers specific foundations coverin g controller workings, feedback control systems and advance process control. At the beginning of the computers, sometime in 1950’s under Cybernetics and Macy Conferences, when modern computer was yet to be born Norbert Weiner was the tallest name in Control Theory. Control Theory progressed well independent of modern computing. Even today control theory remains one of the best hereditary ancestor of Artificial Intelligence. Enough has been said into the AI research about influence of psychology, cognitive biology an...

  Recently I was talking to few executives of a Process Plant( Fertilizer company) where they have been contemplating acquisition of a high end engineering solution which maps in detail all the mass and energy balances in a fertilizer plant. Objective of this system was to improve plant’s overall efficiency, bring about dramatic changes in understanding of material consumption, plant’s blind spots and take plant to the next frontier of efficiency. Normally most of the process plants carry out such initiatives to improve overall plant capacity, also known as debottlenecking. This debott l enecking yields good results if done properly over and over for few times. Next level is to see energy consumption optimization which again like de-bottlenecking is a continuous journey. Further to this process plants try to employ systems which have strong engineering model of plant capturing mass and energy balance equations to stat with and other next level of engineering equation...

  Technology in all areas is viewed as set of inanimate objects. Simondon’s concept of “Individuation” really transforms the perspective to a refreshingly new area where systems, components, devices and pieces of technology in all aspects of life seem to have their own process of “Individuation”. While this process in interesting the present times are making systems more autonomous and more thinking oriented. The the next version of such mass “Thi n king on its own orientation” is Artificial Intelligence and more so Machine Learning. This coupled with IoT and Industry 4.0 also gives interesting paradigm of “Machine/Device/System-As-A-Service” Under “As-A-service there is definite catalog of services available and you can consume them, other machines/devices/systems can also consume them. This opens interesting possibilities like “Social networks of the In-animate, individuated objects”. From Simondon’s “Technical Objects” to Yuk Hui’s “Digital Objects” journey moves t...

 Area Business Use Case Technique/Algorithm Business Impact Fraud detection Insurance fraud brings va s t financial loss to insurance companies every year. Data science platforms make it possible to detect fraudulent activity, suspicious links, and subtle behavior patterns using multiple techniques. To make this detection possible the algorithm should be fed with a constant flow of data. Usually, insurance companies use statistical models for efficient fraud detection. These models rely on the previous cases of fraudulent activity and apply sampling method to analyze them. In addition, predictive modeling techniques are applied here, for the analysis and filtering of fraud instances. Identifying links between suspicious activities helps to recognize fraud schemes that were not noticed before. Decision Trees, Random Forests, Restricted Boltzmann Machines Reduction of Fraud translates into better outcomes as well as better bottom-line Price Optimization Price optimizatio...

  For years domain specialists have been talking against the pure statistical approach to solve proble m s and the same is now happening with statisticians as compared to AI/ML brute force nerds. Gradually this was to happen. Pure play domain specialists who would build models from physics, thermodynamics and engineering disciplines have been doing model building for years. Domain specialists eventually realised that they cannot math model all situations, some problems they realised that are almost intractable with the 19980’s model of computing. This is where statisticians with few techniques like ANONA, Multivariate analytics, regression and principal component analysis conquered the domain without knowing the domain. New age AI/ML ( I am not talking about GOFAI, good old friendly AI) too is as of now seen by both domain and stats guys as brute-force. Well not all of that is brute-force. But yet for business decision makers to start looking at results of AI/ML and st...

  Rise of Creativity because of Rise of Machines Industrialization started only around 250 years ago. The focus of industrialization was to reduce human efforts.In the same way focus of AI/ML is to reduce human thinking. This way, whatever is mechanical and can be automated will be automated in next 50 years ( or earlier). What it means to humanity is, everybody will have lot of time at the disposal to do creative work. To do work which AI/ML algorithms cannot do. This will leave a lot of time to pursue creative endeavors. This is the big positive impact of AI/ML Society in next few years will be dramatically reorganized. It will be a cybernetic-human society. Machine will do lot of tasks done by human and human will be free to do lot of new tasks, mostly creative tasks. Countries and political systems need to be reorganized too. They need to ensure that minimum income, health and education to all. Ideally to survive work( as we know work today) should not be required. What we...

          Author with her immaculate background with Harvard and combination of Philosophy and Sociology education brings out unique perspectives about new kind of capitalism, Surveillance Capitalism. Technology companies have created huge impact on human lives in last 100 years. The Big Tech companies impact everyday life in more and more ways in our everyday life. Author brings out unique concepts like "Behavioral Surplus" and "Instrumenterian Power" to mae her point. The strong theoretical basis is brought by bringing out original ideas of Max Weber, Karl Marx and Emily Drukheim.      First, a brief description of surveillance capitalism. It “unilaterally claims human experience as free raw material for translation into behavioral data. Although some of these data are applied to product or service improvement, the rest are declared as a proprietary behavioral surplus, fed into advanced manufacturing processes known as ‘machine intellig...