GROUP D13-3 (INSTRUMENTATION)

WE WILL BE PRESENTING THE FOLLOWING:

• INTRODUCTION TO MACHINE LEARNING

• THE BASICS OF MACHINE LEARNING

• APPLICATIONS OF MACHINE LEARNING IN INDUSTRYo PRODUCT CATEGORIZATIONo IMPROVING ACCURACY OF INERTIAL

MEASUREMENT UNIT USING SUPERVISED MACHINE LEARNING

o DATA MINING TECHNIQUESo MACHINE LEARNING FOR MEDICAL DIAGNOSIS

• FUTURE SCOPE OF MACHINE LEARNING

MACHINE LEARNING

DEFINITION OF MACHINE LEARNING

Ability of a machine to improve its own performance through the use of a software that employs artificial intelligence techniques to mimic the ways by which humans seem to learn such as repetition and experience.

Helps in building machines exhibiting intelligent behavior.

Apart from artificial intelligence it is also used in administration, commerce and industry.

The most widely known demonstration of this migration is ‘DATA MINING’.

BENEFITS:Makes human-computer

interaction easierRelatively simple to integrateWill distinguish your products from

othersIncrease customer satisfactionWill improve simple-intelligent

systems

APPLICATIONS:Medical diagnosisData miningBioinformaticsSpeech and handwriting recognitionProduct categorizationInertial measurement unit (IMU)Information retrieval

CONCEPT OF MACHINE LEARNING

What exactly is“Machine Learning”??

Input Output

(sensors) (Predictions)

BLACKBOX

Database+

Set of Rules

• Machine learning, a branch of artificial intelligence, is a scientific discipline concerned with the design and development of algorithms.

• Computer system (expert system) which is imbued with decision making ability like a human expert.

• Two parts: a. Knowledge base (database) b. Inference engine (predicted with certain

probability)• It is a learner (like a small baby) which looks at the

examples (obstacles), analyses it using stored data which also includes previous experiences, finds its algorithm and predicts the possible solution with highest probability.

• It keeps updating itself with every obstacle solved, enhancing its performance every time.

• Its algorithms includes different combinations of logic.

• Number of types of obstacles in real world are huge, hence it has to proceed to a generalize solution using certain set of rules.

• Only disadvantage being its high error-making.

• And thus human interfacing with these systems becomes necessary.

Necessity of Human Machine Interfacing

So How Does These Expert Systems Differ from Humans...??

Product Categorization

Need for Product Categorization

Difference in the Scale of Companies

So How Does it Work ...???

Features of Product Categorization

• Classification

• Filtering

• Suggestions / history

• User personal information

IMU SYSTEM USING SUPERVISED MACHINE

LEARNING

ANGLE MEASUREMENT

Motion Capture Technology

Disadvantages of Motion Capture

• Specific hardware and special software increase cost

• Camera field of view is necessary

• More space required

• No calibrations or manipulation while recording data

• Sometimes we require more than one camera for accuracy

• needs proper lightning conditions

So What Should We Do...??

Inertial Measurement Unit (IMU) system with Machine Learning

Inertial Measurement Unit (IMU)with Machine Learning

• IMU is sensor which measures acceleration and angular velocity rate.

• The inertial measurement unit using support vector regression method has the advantages of having a small size as well as quite low cost.

• As compared to the motion capture system, it provides better positional data analysis.

Enhancement in Accuracy of IMU

Kernel trick

-Inner product space

-Linear analysis

Support vector machine

-Classification

-Regression analysis

Data Mining

What is Data Mining?

• Intersection of computer science and statistics

• Data mining software is analytical tools for analyzing data.

• Data mining is process of finding correlations or patterns among large databases.

What is a 'Pattern'?

It is the probability of distribution of similar data. Or in other words its just a relation between the variables.

Machine learning and Data Mining:

1. The computer sorts the data based

on the algorithm.

2. If there is some drastic change in data then,

the algorithm tries to find relation between

them and adapts accordingly.

Identifying non-trivial, valid and useful patterns in a given database is known as 'Knowledge Discovery in Databases (KDD)'.

Steps:• Understand and define problem.• Extract Data:

We should extract data what we need from the Database.• Data Engineering:

Deal with missing variables, rescale data, Combine similar attributes.• Algorithm Engineering (This is the ML part):

Figure out what algorithm to use or write one.

An Overview of the Knowledge Discovery in Database (KDD) Process

Applications :

• DM for Artificial Neural Networks : In most cases a neural network is an adaptive system

that changes its structure so Data Mining is used to model complex relationships between inputs and outputs or to find patterns in data.

• Instance-based Learning Algorithms for

DM : Instance-Based Learning (IBL) is defined as the

generalizing of a new instance to be classified from the stored training examples, which is widely used for classification tasks.Here actually the machine learn from the experience.

Machine learning for Medical Diagnosis

Me Medical Diagnosis by Machine Learning

• Medical diagnosis: It is a procedure to identify disorder in a person.

• Machine learning for medical diagnosis:

It means that the computer will identify the symptoms and tell what that particular person is diagnosed with.

SURVEY OF DIAGNOSIS BY ALGORITHMS/ PHYSICIANS

SELECTION OF THE APPROPRIATE MACHINE LEARNING SYSTEM

Good performanceThe transparency of diagnostic

knowledge. The ability to explain decisions The ability of the algorithm to reduce

the number of tests necessary to obtain reliable diagnosis.

The ability to appropriately deal with missing data.

MACHINE LEARNING ALGORITHMS

1. STATISTICAL OR PATTERN RECOGNITION.

2. INDUCTIVE LEARNING OF SYMBOLIC RULES.

3. ARTIFICIAL NEURAL NETWORKS.

Medical Imaging:

• CCD and GDV are types of image devices which have found great applications in Machine Learning Systems.

• Medical Imaging is taking photos of body parts (both internal and external) and analyzing them for a disorder.

CONCLUSIONIN FUTURE, THE STUDY OF MACHINE

LEARNING HOLDS EXCITING PROSPECTS WITH CONSTANT INNOVATIONS IN DIVERSE FIELDS.

WITH BETTER ALGORITHMS, WE CAN COMPLETELY BRIDGE THE GAP BETWEEN MEN AND MACHINES.

BECAUSE IT IS A TYPE OF ADAPTIVE LEARNING, IT WILL FIND APPLICATIONS IN ALL POSSIBLE FIELDS.

THANK YOU!