100% found this document useful (1 vote)
56 views72 pages

Research Areas in Artificial Intelligence and Machine Learning

Research areas in Artificial Intelligence and Machine Learning

Uploaded by

MM PrasadaReddy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
56 views72 pages

Research Areas in Artificial Intelligence and Machine Learning

Research areas in Artificial Intelligence and Machine Learning

Uploaded by

MM PrasadaReddy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 72

RECENT TRENDS IN SIGNAL AND

IMAGE PROCESSING
AI & ML
RESEARCH AREAS

Prof. S.Varadarajan
Professor, Dept. of ECE
S.V.University, Tirupati.
,

Sapience means Sagacity/wisdom; Homo-Sapiens means Humans


Latest Applications in AI
• Computer Based Weather forecast: Feb’2020
In the past, we have seen a constant rate of acceleration in information processing power as
predicted by Moore's Law, but it now looks as if this exponential rate of growth is limited. New
developments rely on artificial intelligence and machine learning

Breaking down complex systems into individual components

SPA or Scalable Probabilistic Approximation is a mathematically-based concept. The method


could be useful in various situations that require large volumes of data to be processed
automatically, such as in biology, for example, when a large number of cells need to be classified
and grouped. "What is particularly useful about the result is that we can then get an
understanding of what characteristics were used to sort the cells
Another potential area of application is neuroscience. Automated analysis of EEG signals could
form the basis for assessments of cerebral status. It could even be used in breast cancer diagnosis,
as mammography images could be analyzed to predict the results of a possible biopsy.
This method enables us to carry out tasks on a standard PC that previously would have required a
supercomputer
Shape Changing Soft robots: Stanford University researchers-March-2020
Soft Robotics is the specific subfield of robotics dealing with constructing robots from highly compliant materials, similar to
those found in living organisms.

• "A significant limitation of most soft robots is that they have to be attached to a
bulky air compressor or plugged into a wall, which prevents them from moving,".
From outer space to your living room
• Their safe-but-sturdy softness may make them useful in homes and
workplaces, where traditional robots could cause injury.
• "This robot could be really useful for space exploration -- especially
because it can be transported in a small package and then operates
untethered after it inflates," "On another planet, it could use its
shape-changing ability to traverse complicated environments,
squeezing through tight spaces and spreading over obstacles."
• CHANDRAYAN – 3 ( Chandryan -2 orbiter,
Vikram-Lander, Pragyan-Rover )
Researchers restore injured man's sense of touch using brain-computer interface
technology, Ohio State Univ.
Date:April 23, 2020

• Researchers have been able to restore sensation to the hand of a research


participant with a severe spinal cord injury using a Brain-Computer interface
(BCI) system. The technology harnesses neural signals that are so minuscule
they can't be perceived and enhances them via artificial sensory feedback sent
back to the participant, resulting in greatly enriched motor function.
• The advances in the BCI system led to three important improvements.
• They enable Burkhart to reliably detect something by touch alone: in the
future, this may be used to find and pick up an object without being able to see
it.
• The system also is the first BCI that allows for restoration of movement and
touch at once, and this ability to experience enhanced touch during movement
gives him a greater sense of control and lets him to do things more quickly.
• Finally, these improvements allow the BCI system to sense how much pressure
to use when handling an object or picking something up -- for example, using a
light touch when picking up a fragile object like a Styrofoam cup but a firmer
grip when picking up something heavy.
Keywords -AI ML, DL, DS

• AI-Enables the Machine to Think-without human


intervention ( using An application of DL,ML )
• Eg: Self driving car
• ML- subset of AI- provides us statistical tools to explore (
to understand)and analyse the data.
Approaches: Supervised learning( Past labelled data) Eg:
Height, Weight for deciding whether the person is Obese or Fit
Eg: Class Room Teaching

• Unsupervised : don’t have any past data clustering data


say using Euclidean distance,
DEEP LEARNING(DL)
• DL:Subset of ML-Machines to learn on its
own- to mimic human brain- Uses Multi
Neuron N/w architecture- DL Neural NWs.

• Various Techniques used for DL :


ANN- data in the form of Nos ;
CNN- i/p in the form of images:
RNN-If the i/p is in the form of Time series
data
ML and DL – derive AI
DATA SCIENCE/DATA SCIENTIST

• DS: ML, DL and Mathematical tools like,


Probability, Statistics, linear Algebra,
Differential calculus.
• Data Scientist may work in ML, DL
• Data Scientist: any one Progr.Lang.- Python/R,
ML,IDE,(spyder), Mathematics- Linear algebra,
diff calculus, Statistics, Data Visualisation
Self-driving cars: Reinforcement learning is used in self-
driving cars for various purposes such as lane changing,
Parking, the following. Amazon cloud service such as
DeepRacer can be used to test RL on physical tracks.

Personalised video recommendations based on different factors related to


every individual.
• AI-powered stock buying/selling: While supervised learning
algorithms can be used to predict the stock prices, it s the
reinforcement learning which can be used to decide whether
to buy, sell or hold the stock at given predicted price.

• RL can be used for NLP use cases such as text summarization,


question & answers, machine translation.

• RL in healthcare can be used to recommend different


treatment options. While supervised learning models can be
used to predict whether a person is suffering from a disease
or not, RL can be used to predict treatment options given a
person is suffering from a particular disease.
ML-APPLICATIONS
• ML Examples available in MATLAB
Nest Cam IQ is intelligent and can tell the difference between people and
things and can automatically follow a person's movements
THE HEART OF AI IS
ANN.
a) The part of the neuron which helps in the acquisition of information is
known as the dendrite. They are tree-like structures that are designed to
receive communications from other cells.
b) The part of the neuron through which information travels as an electrical
impulse is known as the axon of the neuron. Axons mostly are covered with
Myelin Sheath which increases the speed of signal transmission
PERCEPTRON
Although today the Perceptron is widely recognized as an
algorithm, it was initially intended as an image recognition
machine. It gets its name from performing the human-like
function of perception, seeing and recognizing images.
• inputs are combined in a weighted sum and, if the weighted sum exceeds a
predefined threshold, the neuron fires and produces an output..
• Threshold T represents the activation function. If the weighted sum of the
inputs is greater than zero the neuron outputs the value 1, otherwise the
output value is zero.

• The activation function decides whether a neuron should be activated or not


by calculating the weighted sum and further adding bias to it. The purpose of
the activation function is to introduce non-linearity into the output of a
neuron.
• BIAS
• Weights control the signal (or the strength of the connection) between two
neurons. In other words, a weight decides how much influence the input will
have on the output. Biases, which are constant, are an additional input into
the next layer that will always have the value of 1.

• What is bias in a neural network? In simple words, neural network bias can
be defined as the constant which is added to the product of features and
weights. It is used to offset the result. It helps the models to shift the
activation function towards the positive or negative side.
The hidden layer is between the input layer and the output layer. It takes in
a set of weighted inputs and produces output through an activation function.
This layer is named hidden because it does not constitute the input or the
output layer. This is the layer where all the processing happens. For
example, a hidden layer functions that are used to identify human eyes and
ears may be used in conjunction by subsequent layers to identify faces in
images. While the functions to identify eyes alone are not enough to
independently recognize objects, they can function jointly within a neural
network.
Cost Function

is a function that measures the performance of a Machine


Learning model for given data. Cost Function quantifies the error
between predicted values and expected values and presents it in
the form of a single real number. Depending on the problem Cost
Function can be formed in many different ways.
• What is a loss/Cost function? ‘Loss’ in Machine learning helps us
understand the difference between the predicted value & the
actual value. The Function used to quantify this loss during the
training phase in the form of a single real number is known as
“Loss Function”. These are used in those supervised learning
algorithms that use optimization techniques. Notable examples
of such algorithms are regression, logistic regression, etc. The
terms cost function & loss function are analogous.

• Loss function: Used when we refer to the error for a single


training example.
• Cost function: Used to refer to an average of the loss functions
over an entire training dataset.
• Why on earth do we need a cost function? Consider a scenario where we
wish to classify data. Suppose we have the height & weight details of some
cats & dogs. Let us use these 2 features to classify them correctly. If we plot
these records, we get the following scatterplot:

COST FUNCTION SCATTER PLOT


Blue dots are cats & red dots are dogs
Fig: Probable solutions to our classification problem
• Essentially all three classifiers have very high accuracy but the third
solution is the best because it does not misclassify any point. The reason
why it classifies all the points perfectly is that the line is almost exactly in
between the two groups, and not closer to any one of the groups. This is
where the concept of cost function comes in. Cost function helps us reach
the optimal solution. The cost function is the technique of evaluating “the
performance of our algorithm/model”.
• It takes both predicted outputs by the model and actual outputs and
calculates how much wrong the model was in its prediction.
Types of the cost function
• There are many cost functions in machine learning and each has its use
cases depending on whether it is a regression problem or classification
problem.
• Regression cost Function
• Binary Classification cost Functions
• Multi-class Classification cost Functions

You might also like