Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - priankaswe

Pages: [1]
1
In 1971, Terry Winograd wrote the SHRDLU program while completing his PhD at MIT. SHRDLU features a world of toy blocks where the computer translates human commands into physical actions, such as “move the red pyramid next to the blue cube.” To succeed in such tasks, the computer must build up semantic knowledge iteratively, a process Winograd discovered was brittle and limited.

The rise of chatbots and voice activated technologies has renewed fervor in natural language processing (NLP) and natural language understanding (NLU) techniques that can produce satisfying human-computer dialogs. Unfortunately, academic breakthroughs have not yet translated to improved user experiences, with Gizmodo writer Darren Orf declaring Messenger chatbots “frustrating and useless” and Facebook admitting a 70% failure rate for their highly anticipated conversational assistant M.

Nevertheless, researchers forge ahead with new plans of attack, occasionally revisiting the same tactics and principles Winograd tried in the 70s. OpenAI recently leveraged reinforcement learning to teach to agents to design their own language by “dropping them into a set of simple worlds, giving them the ability to communicate, and then giving them goals that can be best achieved by communicating with other agents.” The agents independently developed a simple “grounded” language.

 MIT Media Lab presents this satisfying clarification on what “grounded” means in the context of language: “Language is grounded in experience. Unlike dictionaries which define words in terms of other words, humans understand many basic words in terms of associations with sensory-motor experiences. People must interact physically with their world to grasp the essence of words like “red,” “heavy,” and “above.” Abstract words are acquired only in relation to more concretely grounded terms. Grounding is thus a fundamental aspect of spoken language, which enables humans to acquire and to use words and sentences in context.”

The antithesis of grounded language is inferred language. Inferred language derives meaning from words themselves rather than what they represent. When trained only on large corpuses of text, but not on real-world representations, statistical methods for NLP and NLU lack true understanding of what words mean. OpenAI points out that such approaches share the weaknesses revealed by John Searle’s famous Chinese Room thought experiment. Equipped with a universal dictionary to map all possible Chinese input sentences to Chinese output sentences, anyone can perform a brute force lookup and produce conversationally acceptable answers without understanding what they’re actually saying.

 

LANGUAGE COMPLEXITY INSPIRES MANY NATURAL LANGUAGE PROCESSING (NLP) TECHNIQUES
Percy Liang, a Stanford CS professor and NLP expert, breaks down the various approaches to NLP / NLU into four distinct categories:

1) Distributional
2) Frame-based
3) Model-theoretical
4) Interactive learning

You might appreciate a brief linguistics lesson before we continue on to define and describe those categories. There are three levels of linguistic analysis:

1) Syntax – what is grammatical?
2) Semantics – what is the meaning?
3) Pragmatics – what is the purpose or goal?

Source: https://www.topbots.com/4-different-approaches-natural-language-processing-understanding/

2
Software Engineering / What is Numerical Analysis
« on: September 16, 2018, 03:31:10 PM »
Numerical analysis, area of mathematics and computer science that creates, analyzes, and implements algorithms for obtaining numerical solutions to problems involving continuous variables. Such problems arise throughout the natural sciences, social sciences, engineering, medicine, and business. Since the mid 20th century, the growth in power and availability of digital computers has led to an increasing use of realistic mathematical models in science and engineering, and numerical analysis of increasing sophistication is needed to solve these more detailed models of the world. The formal academic area of numerical analysis ranges from quite theoretical mathematical studies to computer science issues.

With the increasing availability of computers, the new discipline of scientific computing, or computational science, emerged during the 1980s and 1990s. The discipline combines numerical analysis, symbolic mathematical computations, computer graphics, and other areas of computer science to make it easier to set up, solve, and interpret complicated mathematical models of the real world.

Source: https://www.britannica.com/science/numerical-analysis

3
Data scientists use many different kinds of machine learning algorithms to discover patterns in big data that lead to actionable insights. At a high level, these different algorithms can be classified into two groups based on the way they “learn” about data to make predictions: supervised and unsupervised learning.
Supervised machine learning is the more commonly used between the two. It includes such algorithms as linear and logistic regression, multi-class classification, and support vector machines. Supervised learning is so named because the data scientist acts as a guide to teach the algorithm what conclusions it should come up with. It’s similar to the way a child might learn arithmetic from a teacher. Supervised learning requires that the algorithm’s possible outputs are already known and that the data used to train the algorithm is already labeled with correct answers. For example, a classification algorithm will learn to identify animals after being trained on a dataset of images that are properly labeled with the species of the animal and some identifying characteristics.
On the other hand, unsupervised machine learning is more closely aligned with what some call true artificial intelligence — the idea that a computer can learn to identify complex processes and patterns without a human to provide guidance along the way. Although unsupervised learning is prohibitively complex for some simpler enterprise use cases, it opens the doors to solving problems that humans normally would not tackle. Some examples of unsupervised machine learning algorithms include k-means clustering, principal and independent component analysis, and association rules.
While a supervised classification algorithm learns to ascribe inputted labels to images of animals, its unsupervised counterpart will look at inherent similarities between the images and separate them into groups accordingly, assigning its own new label to each group. In a practical example, this type of algorithm is useful for customer segmentation because it will return groups based on parameters that a human may not consider due to pre-existing biases about the company’s demographic.
Choosing to use either a supervised or unsupervised machine learning algorithm typically depends on factors related to the structure and volume of your data and the use case of the issue at hand. A well-rounded data science program will use both types of algorithms to build predictive data models that help stakeholders make decisions across a variety of business challenges.

Source: https://www.datascience.com/blog/supervised-and-unsupervised-machine-learning-algorithms

4
Software Engineering / WHY DO WE HAVE TO STUDY NUMERICAL ANALYSIS?
« on: September 16, 2018, 10:50:59 AM »
In the teaching of mathematics in engineering careers, the importance and usefulness of the different issues studied must be emphasized to capture the interest of students. The discussion of theory within the context of problem solving, not only motivates students but also helps them to achieve significant and comprehensive learning. This methodology may be useful to solve the disruption between mathematics and specialty subjects.

Some experiences have been carried out in Numerical Analysis courses at Facultad Regional San Nicolás, Universidad Tecnológica Nacional in Argentina. A great motivation in students has been observed when working with visual applications dealing with different issues in the subject, designed using different Computer Algebra Systems (CAS).

With the aim that students achieve meaningful learning, it was decided to include problems where numerical methods are necessary to obtain their solution. So as students can focus on the problems and the solution techniques needed, without wasting time making calculations, some apps were designed to be used for solving them.

The newest applications were designed with Mathematica, as computable document format files (CDF). The program required to open these files is the CDF-Player, available for free in the Wolfram´s website. Also, this kind of files may be embedded in a webpage, and they can be executed in a web browser if the corresponding plugin is installed. In this way, students can access to the CDF files almost everywhere.

Source: https://library.iated.org/view/CALIGARIS2016STU



5
Nasa has named the astronauts who will fly the first missions into space on commercially provided rockets and capsules, starting next year.

The nine individuals will go up on systems developed by - and contracted from - the Boeing and SpaceX firms.

Most have previous experience in orbit. Among them are the commander and pilot of the final shuttle mission in 2011.

For the past seven years, Russian rockets have been the only way to get people into orbit.

Nasa Administrator Jim Bridenstine introduced the astronauts during a ceremony at the Johnson Space Center in Houston, Texas.

"For the first time since 2011, we are on the brink of launching American astronauts on American rockets from American soil," he told the enthusiastic audience.

The commander on the historic last shuttle mission, Chris Ferguson, is now a Boeing employee and has been heavily involved in developing the company's CST-100 Starliner capsule.

When this ship makes its maiden crewed flight in the middle of next year, launching atop an Atlas rocket from Cape Canaveral, Ferguson will be joined by Eric Boe and Nicole Aunapu Mann.

Boe is a former shuttle pilot; Mann will be making her first trip into space.

The SpaceX Dragon capsule, on current timelines, is set to make its maiden crewed flight in April. It will ride atop a Falcon-9 rocket from the Kennedy Space Center. It will actually use one of the old shuttle pads, although this has now been modified for the smaller Falcon.

At the helm will be Doug Hurley and Bob Behnken. Hurley was the pilot on the last shuttle mission. Behnken has been in space on two previous occasions.

The initial crewed flights by Boeing and SpaceX will spend a short period in orbit - measured perhaps in days or a few weeks, and attached to the International Space Station (ISS) - before coming back to Earth.

It is on later missions that the crew capsules will go to the station and dock for more extended stays, and Nasa also named the astronauts for those first flights as well.

For Boeing, this introductory long-duration mission includes Josh Cassada, who has never been in space before, and Suni Williams, who is one of the most experienced American astronauts in history, having spent a cumulative 321 days in orbit through her career.

For SpaceX, such an ISS mission would involve Victor Glover, another newbie, and Mike Hopkins, who has already spent 166 days on the sky-high lab during two tours of duty.

Nasa took the decision after the shuttle retired to turn transportation to low-Earth obit destinations into a service that it could buy. It has given seed money to SpaceX and Boeing to incentivise them, but the companies themselves have also had to invest their own money.

Nasa's motivation was to save money it could then spend on a rocket and capsule system to take humans back to the Moon and on to Mars.

That system will likely fly some time early next decade.

Source link: https://www.bbc.com/news/science-environment-45060553

6
Discrete mathematics is the branch of mathematics dealing with objects that can assume only distinct, separated values. The term "Discrete Mathematics" is therefore used in contrast with "Continuous Mathematics," which is the branch of mathematics dealing with objects that can vary smoothly (and which includes, for example, calculus). Whereas discrete objects can often be characterized by integers, continuous objects require real numbers.

The study of how discrete objects combine with one another and the probabilities of various outcomes is known as combinatorics. Other fields of mathematics that are considered to be part of discrete mathematics include graph theory and the theory of computation. Topics in number theory such as congruence’s and recurrence relations are also considered part of discrete mathematics.

The study of topics in discrete mathematics usually includes the study of algorithms, their implementations, and efficiencies. Discrete mathematics is the mathematical language of computer science, and as such, its importance has increased dramatically in recent decades.

The set of objects studied in discrete mathematics can be finite or infinite. The term finite mathematics is sometimes applied to parts of the field of discrete mathematics that deals with finite sets, particularly those areas relevant to business.

Research in discrete mathematics increased in the latter half of the twentieth century partly due to the development of digital computers which operate in discrete steps and store data in discrete bits. Concepts and notations from discrete mathematics are useful in studying and describing objects and problems in branches of computer science, such as computer algorithms, programming languages, cryptography, automated theorem proving, and software development. Conversely, computer implementations are significant in applying ideas from discrete mathematics to real-world problems, such as in operations research.


IMPORTANCE OF DISCRETE MATHEMATICS IN COMPUTER SCIENCE
Achieving working knowledge of many principles of computer science requires mastery of certain relevant mathematical concepts and skills. For example, A grasp of Boolean algebra including DeMorgans Law is useful for understanding Boolean expressions and the basics of combinational circuits concepts surrounding the growth of functions and summations are useful for analysis of loop control structures exposure to solving recurrence relations is de rigeur for the analysis of recursive algorithms and an introduction to proof methods facilitates consideration of program correctness and thinking rigorously in general.

Students are introduced to proof techniques before they begin to consider the idea of proving programs correct. They learn about propositional logic and Boolean algebra before they study some very elementary circuits and learn decision control structures and Boolean variables. They are introduced to predicate logic near the time they are beginning programming and learning about variables. They learn about growth of functions big-O notation and summations before they analyze loops and nested loops and they have the tools to begin algorithm analysis from the time they first begin to learn about iterative constructs. In conjunction with an introduction to number theory they do laboratory and programming exercises involving an assortment of integer algorithms.

Students learn about recursive definitions recurrence relations, analyzing recursive algorithms and writing recursive algorithms and programs together in the same course. They study matrices and matrix manipulations in conjunction with the array data structure. They learn about permutations and combinations, relations, graphs, and trees at the same time that their programming knowledge and sophistication are improving and they can do increasingly interesting programming exercises involving these concepts.


Source link: http://cybercomputing.blogspot.com/2012/06/discrete-mathematics-applications-and.html

7
Abstract: Spell checkers check whether a word is
misspelled and provide suggestions to correct it.
Detection and correction of spelling errors in Bangla
language which is the seventh most spoken native
language in the world, is very onerous because of the
complex rules of Bangla spelling. There is no systematic
literature review on this research topic. In this paper, we
present a systematic literature review on checking and
correcting spelling errors in Bangla language. We
investigate the current methods used for spell checking
and find out what challenges are addressed by those
methods. We also report the limitations of those methods.
Recent relevant studies are selected based on a set of
significant criteria. Our results indicate that there are
research gaps in this research topic and has a potential for
further investigation

Full paper link: http://www.mecs-press.org/ijmecs/ijmecs-v9-n6/IJMECS-V9-N6-6.pdf

8
Natural language processing​ or NLP is a branch of artificial intelligence that has many important implications on the ways that computers and humans interact. Human language, developed over thousands and thousands of years, has become a nuanced form of communication that carries a wealth of information that often transcends the words alone. NLP will become an important technology in bridging the gap between human communication and digital data. Here are 5 ways that natural language processing will be used in the years to come.
1. Machine Translation
2. Fighting Spam
3. Information Extraction
4. Summarization
5. Question Answering


Source Link: https://www.lifewire.com/applications-of-natural-language-processing-technology-2495544

9
Software Engineering / Everyday applications of discrete mathematics
« on: July 04, 2018, 02:56:33 PM »
Computers run software and store files. The software and files are both stored as huge strings of 1s and 0s. Binary math is discrete mathematics.
Networks are, at the base, discrete structures. The routers that run the internet are connected by long cables. People are connected to each other by social media ("following" on Twitter, "friending" on Facebook, etc.). The US highway system connects cities with roads.
Doing web searches in multiple languages at once, and returning a summary, uses linear algebra.
Google Maps uses discrete mathematics to determine fastest driving routes and times. There is a simpler version that works with small maps and technicalities involved in adapting to large maps.
Scheduling problems---like deciding which nurses should work which shifts, or which airline pilots should be flying which routes, or scheduling rooms for an event, or deciding timeslots for committee meetings, or which chemicals can be stored in which parts of a warehouse---are solved either using graph coloring or using combinatorial optimization, both parts of discrete mathematics. One example is scheduling games for a professional sports league.
An analog clock has gears inside, and the sizes/teeth needed for correct timekeeping are determined using discrete math.
Machine Job Scheduling: Scheduling tasks to be completed by a single machine uses graph theory. Scheduling tasks to be completed by a set of machines is a bin-packing problem, which is part of discrete optimization. Google describes the issue for multiple types of jobs on multiple machines.
Cell phone communications: Making efficient use of the broadcast spectrum for mobile phones uses linear algebra and information theory. Assigning frequencies so that there is no interference with nearby phones can use graph theory or can use discrete optimization.
Digital image processing uses discrete mathematics to merge images or apply filters.

Source: http://www.mathily.org/dm-rw.html

10
Software Engineering / Taming fusion with machine learning
« on: July 04, 2018, 02:19:06 PM »
“In some ways you can compare what we are trying to do to self-driving cars.”
MIT postdoc Cristina Rea is describing her work at the Institute's Plasma Science and Fusion Center (PSFC), where she is exploring ways to predict disruptions in the turbulent plasma that fuels fusion tokamak reactors..
“You want to be able to predict when an object presents an obstacle for your car,” she continues with her comparison. “Likewise, in fusion devices, you need to be able to predict disruptions, with enough warning time so you can take actions to avoid a problem.”
Tokamaks use magnetic fields to contain hot plasma in a donut-shaped vacuum chamber long enough for fusion to occur. Chaotic and unpredictable, the plasma resists confinement, and disrupts. Timely predictions about incipient plasma disruptions could help sustain fusion energy production in these devices, while preventing damage to the machine.
To tackle the issue, Rea is part of the PSFC’s collaboration with the DIII-D tokamak in San Diego, which, since the Center’s Alcator C-Mod device ended its run in September 2016, is the only fusion-grade tokamak in the U.S. that is currently running. With PSFC research scientist Bob Granetz, she is developing a database to centralize information about disruptions in C-Mod, DIII-D, and other tokamaks around the world. Employing machine learning techniques to model how disruptions are likely to progress, she hopes to use these data to find ways to mitigate the problem.
Her interest in plasma science developed at the University of Padua, where as a graduate student she was able to gain experience doing research at Italy’s Reversed Field Pinch fusion experiment. She graduated with a PhD in physics in 2015. So it felt like a detour when she found herself accepting a position as a data scientist with one of Italy’s largest banks.
“UniCredit was looking for a PhD that could learn fast, and so they let me have some hands-on training in machine learning techniques, such as the ones used to improve the customer relationship management. We used bank customers’ account and credit card data to predict the probability that a particular customer would close the account and abandon the bank.”
Although genuinely intrigued by the specific techniques used to explore and analyze data, she notes, “the subject itself was not what I studied, what I loved.”
She decided, with her husband, who is also a plasma physicist, “to give research another shot.” Together they began exploring opportunities outside of Italy. When her husband received a postdoctoral position at General Atomics (GA) in San Diego, Rea turned her attention towards the local opportunities, including DIII-D, which is housed at GA. There she discovered the perfect project for her talents.
MIT research scientist Bob Granetz, collaborating at DIII-D, needed a postdoc versed in plasma physics to help centralize data about tokamak plasma disruptions, in preparation for analyzing the data further with machine learning.
“I knew the methods,” says Rea, “though in an entirely different context. That was one lucky shot!”
Despite a wealth of knowledge gathered from fusion and plasma devices, no consistent or solid theory of the disruption process has yet been developed.
“Disruptions are complicated phenomena, and we don't have sufficient theoretical understanding to be able to calculate when a disruption is approaching,” says supervisor Granetz.  “Although we have huge amounts of empirical data, it's difficult for us to recognize the necessary disruption-relevant information that we believe is contained in the data. This is exactly the kind of situation that machine learning is tailored for: Take a few algorithms that are good at recognizing patterns in data, train the algorithms on our large databases of disruption-relevant data, and see if something useful comes out.”
Rea is excited to apply some of the machine learning techniques she acquired in banking to this complex problem. Being on site at DIII-D allows Rea to gather data on the major working tokamak in the U.S., and to be ready to test any promising algorithms on the tokamak in real time. Her recent research compares data from DIII-D and C-Mod in order to create an algorithm for predicting disruptions that would be broadly applicable to other tokamaks, including ITER, the next-generation fusion device being built in France, and, SPARC, the PSFC’s compact, high-field, net fusion energy experiment. Her early results have indicated that the chosen algorithm has wildly different predictive success on each device, with the success on C-Mod actually being worse than a random guess.
“DIII-D has had the better results so far in terms of reliable predictions and we were able to develop a robust model, given the experimental data available. C-Mod has shown completely different disruption dynamics, and we are actively working to develop an algorithm capable of both capturing these dynamics and at the same time reliably generalizing to unseen cases.”
Rea is more energized than deterred by the challenge of developing cutting-edge machine learning tools that will eventually provide a process for consistent predictions. She has always loved a challenge, deciding in college on a physics major instead of English literature because it posed more questions that she wanted to answer.
“In college I was first interested in physics from a theoretical point of view. Then I switched to plasma. Every time I switched to a different topic it was because I needed more questions. I needed more problems.” She laughs as she questions, “What’s going to really stump me?”
Fusion research will keep the questions coming.
“Plasma physics, and all the efforts going into understanding and trying to develop fusion energy — I think this is really a good task and a good question to solve. Finding out the answer to that question — how to create fusion energy on Earth — will have a major impact, not just on a limited number of people, but for all humankind.”

Source: http://news.mit.edu/2018/taming-fusion-with-machine-learning-cristina-rea-0702

Pages: [1]