0% found this document useful (0 votes)
34 views10 pages

Evaluation

Model evaluation is an essential part of evaluating how well a model will perform on unseen future data. There are different evaluation techniques depending on the model type and purpose. The testing data should not be the same as the training data, otherwise the model will just remember the training data and will not provide accurate predictions. Confusion matrices are used to evaluate models using metrics like accuracy, precision, recall, and F1 score. These metrics help determine if a model correctly identifies positive and negative examples.

Uploaded by

Nikita Agrawal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
34 views10 pages

Evaluation

Model evaluation is an essential part of evaluating how well a model will perform on unseen future data. There are different evaluation techniques depending on the model type and purpose. The testing data should not be the same as the training data, otherwise the model will just remember the training data and will not provide accurate predictions. Confusion matrices are used to evaluate models using metrics like accuracy, precision, recall, and F1 score. These metrics help determine if a model correctly identifies positive and negative examples.

Uploaded by

Nikita Agrawal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 10

UNIT-VIIEVALUATION

ANTRODUCTION
Model Evaluation is an essential part of the model development process. It helps to choose the
best model that represents our data and how well the chosen model will work in the future.
the stage which will help to decide that model is working well or not. In the evaluation process, o
evaluate the model by feeding the testing data and then comparingthe actual answer.
There can be different Evaluation techniques, depending of the type and purpose of the moe
important thing is that we cannot use the training data to test the model for evaluation. Tne
predict the
because the model will remember the whole training data and will therefore always
correct answer/output for any point in the training set. This process is known as over
1.2 MODEL EVALUATION TERMINOLOGY
In model evaluation, we come across many terms like:
machine.
1. Prediction / Predicted Value: The Prediction is the output which is given by the
2. Reality I ActualValue: The Actual Value represents the actual result of the Al Moue
Tounderstand the process of evaluation,we need to take an example for evaluation.
Case 1: We have trained the model for identifving an orange The image shown was ot an ofa9
Prediction matches with the
The model predicted Yes, which means the fruit is orange. The
Reality. Hence, this condition is termedas True Positive.
Case 2: We have given a lemon and asked the machine whether it is an orange of ot
Therefore. this condition is termed as
case, the machine too has predicted it correctly as a No.
True Negative.
model. But the machine has incorrectly
Case 3: Again, we displayed an image of a lemon to the
predicted that it is an orange. This case is termed as False Positive.
again we displayed an image of a orange and because of which the Reality is Yes
Case 4: Now is a
machine has incorrectly predicted it as a No, which means the machine predicts that it
but the
Negative.
lemon. Therefore, this case becomes False
1.3CONFUSION MATRIX
known as Error matrix and is represented by a table which describe
The confusion matrix is also
between the prediction and reality. The confusion matrixX allows us to
the result of comparison four parameters of the matrix.
understand the prediction results. Let us check the

POSITIVE (TP): Prediction matches the reality (True) and is actually positive.
TRUE
NEAANE(FP): Prediction does not match reality (False) and is actually negative.
FALSE
NEGATIVE (TN): Prediction matches the reality (True) and is actually negative.
TRUE
Prediction does not match reality (False) and is actually positive.
FALSE POSITIVE (FP):
[1]
f. The Emi NEGATIVE
|POSITIVE
FP
TP

POSITIVE FN TN
NEGA TIVE

outcomes or parameters are plotted on the Confusion Matrix as follows


Thesefour
Actual Values

Predicted
Values Positive (1) Negatve (0)

Positlve (1) TP FP

Negative (0) TN

1.4 EVALUATION METHODS


values to evaluate an Al model.
confusion matrices, we need to compute the following
Usingthe are
times the predictions out of all the observations
percentage of
1. Accuracy: This is the determine Accuracy is:
correct. The Formula to
+ TN)
Number of correct predictions (TP + FN)
made (TP + TN + FP
ACCuracy =Total Number of predictions

In Percentage,
(TP +TN)
Number of correct predictions FP + FN)
x 100%

Accuracy = Total Number of predictionsmade


(TP + TN+

desirable predictions turn out to be correct. The


rate at which the
2. Precision: This is theis:
Formula for Precision TP
Precision = TP + FP

In Percentage, TP
Precision = TP+ FP x100%

predictions to the overall number of positive instances


3. Recall: It is a rate of correct positive are:
in the dataset. The Formula for calculating Recall
Predictions actually positive
Recall = TP + FN
Actual positive values in the dataset
ASS X

In Percentage,
TP
Recall = X 100%
TP + FN
1.4.1Which motric is important?
The decision to choose precision or recall as an evaluation parameter completely depends upon
the conditions in which the model has been deployed. Procision is used as a metric when our
objective is to minimize falso positives and Recallis usod when the objoctive is to minimize
false negatives.
4. F1 Score: It is a measure of balance between precision and recall. It is computed as per
the following formula:
Precisionx Recal
F1 = 2 X
Precision + Recall

UNIT-VCOMPUTER VISION (CV)

1.1 WHAT IS COMPUTER VISION?


artificial intelligence. Computer vision trains
The Computer Vision is one of the domain of
the visual world.
computers to interpret and understand the images in
CONCEPTS
1.2 UNDERSTANDING CV
depends on pattern recognition techniques for self-training and
Computer vision primarily recognize
image, a computer vision system must
understanding visual data. In atwo-dimensional textures, colors, sizes and among other things,
as shapes,
the objects with its characteristics such
image.
to provide a description of the
COMPUTER VISION
1.3 APPLICATION OF are:
introduced around 1970s. Some of the main application areas
Computer vision concept was sensor
autonomous vehicles, Computer Vision is used with
1. Self-driving Cars: In map.
vehicles can capture real-time images to create a 3D
technology. The cameras on surroundings while
autonomous vehicles can easily make sense of their
Using this map, objects on the road.
identifyingpeople, cars, and other for
vision plays a very important role in providing security
Computer becoming popular in smart schools
2. Facial Recognition: Facial recognition is also
smart cities and smart homes. taken through Artificial intelligence using computer vision
where attendance of student is
technology. the
proven immensely helpful in skin
3. Medical magif: Computer Vision applications have
brain tumours, cancerous
moles in
medical field, especially in the accurate detection of 3D models. This
These applications convert 20 scanned images into
images and so on. their patients.
better insight into the health of
acquire a
helps medical professionals
(3]
vision.
CASS X of
computer
image,
4. anotherutility input
Googte
uses theSearch by lmage mage
Google searh by analyzes
is
the
diferent
features of

1.4 Comparesteitchnolwithothe
gy database
of Computer
Vision as it
of innages and gives the
result.
search

The COMPUTER
area of VISION TASKS see and
perceive.
Many
what istimes,
Computer humans
we look at Vision deals with objects and its surrounding9 that experienceto
an
determine

The
missing inoomplete
from it. All these task image and use our knowBedge
come under the category of
and
Computer Vision.
following are the main objects
1. Computer Vision tasks related to single and multiple
Classification: Image
belongs to. It is appliedclassification
only for singlebasicalty
objects.means
For exanmple. class the
object
identifyingthere are objects present
what
belonging to mnany classes such as ball, cat, sheep etc. The machine or deep learning
model would
2. deteminethat the animal detected in the image belongs to class cat.
Classification
identifying
+
Localisation:
and locating the object Classification processes of
present in that image. It is applied only for single objects
+ Localisation invoves both
3. Object Detection: It is the
along with their ability to detect or identfy objects in any given image correctly
the fom of three-dimensional
It is applied rectangular
position in the given Image. The 3D image is shown init.
boxes (known as Bounding Boxes) which bound the object within
onty for muitiple objects.
4. Instance Segmentation: Instance Segmentation basicaly
identiy
in the image with their
boundarnies intact. In this case, it's not onlydifferent instances g
about finding objects
an image, but also about creating a mask for each
detected obiect that is as accurate 2s
possible. It is applied only for multiple objects.

Classification Instanca
Classification Object Detection
+ Localization Segmentation

CAT, DOG, DX CAT, DOG, D


CAT CAT

Mue s
Singie object

1.5 IMAGE BASICs images. Wel, you know that computers use
about how computers see the Digital Images thatyou see on
Computer vision is all images.
capture images, which are digital graphic) or by plotting some
to (bitmap
cameras as their eyes joining millions of small dots
created via
screens are either graphic).
mathematical equations (vector
1. Pixel The small dots
element) reptesents thethat create an image ale called pixels
terms of colour and smallest piece of the screen that be pixel (short for picture
(brightness). intensty. Each plxel can be sot to a controlled individually in
can
2. different colour and intensy
Resolution: The number of
measures is dimension of anpixelsoverall
the in an image
image ordetermine
its
resolution measUred in width how big the resolution
size
Image resolution
3. Pixol
Value: (vertical of an
The images stored in a pixels) and height (horizontal pixels).
image is Image
with numerical computer are
intensity of values, which are also called composed of pixels. Pixels
color and 255each pixel. They are integers that Pixel Values. These pixel values are denoted
4. represents white color. range from 0to 255, where 0represents represent the
Grayscale black
without Images: Grayscale images are images which have a
Color or apparent color, The darkest
zero value of pixel. The possible shade is black, whichrange of shades of gray
of color or 255 lightest is the total
value of a pixel, possible shade is white, which is the absence of
brightness levels of the three primaryIntermediate shades of gray are represented total presence
5. RGB
Images: colors. by equal
Colored
All other colors are images are made up of three primary
colors - red, green and blue.
images are stored generated by
in the form ofcombining different intensities- of these colors only. RGB
three different channels
of these channels red, green, and blue. Each
have
intensity of the pixels. values ranging from 0 to 255 where these numbers represent the

UNIT- IV DATA SCIENCE


1.1 INTRODUCTION TO DATA SCIENCE
Data science deals with the extraction of relevant
information from data. It is the analysis and
study of data. It also deals with finding hidden patterns in the data. All the three domains has
different type of data but in data sciences, we deal with all these data for the purpose of analysis.
1.2 APPLICATIONS OF DATA SCIENCE

1. Fraud and Risk Detection: Fraud can be detected by using Data Science monitoring and
analysing of the user activity to find any unusual or malicious pattern. For instance,
abnormal variance in credit card usage or banking transactions can be easily trace so that
appropriate action maybe taken on time.
service provided
2. Internet Search or Autocomplete: Information retrieval is one of the best
Ask, etc., make use of
by search engines. Different search engines such as Google, Bing, of
the queries within a fraction
data science algorithms and give the best results for
seconds.
or
Website Recommendation: While doing online shopping on Amazon and Flipkart
3.
Netflix, you must have seen recommended products or videos that suit your taste.
watching
(5)
CLASS X
This is the
recommernder
activity of users and systern, which fnakas uss data s A r , arayzes tt
4.
Diagnosis recormmendsthe relsvart sst Aiterrs to the users.
of Modical
This can be done by Imaging:
plays an important role in Undor the dornain A data scienA, prosvArve a r a t s
prediction from it. analyzingirnproving the
the patient'health of patierts and rnanaging chronc diseaes
5. Digital s history. recoynizing the patterns and generatirg
digital Markoting and Target
marketing
customers. These
help
companies Advertising:
or The Dat Sciene related applications in
applications
related to their past make useorganizations
of addvertise their produts to the right
and alyorithrris
advertisements
1.3 DATA COLLECTION
search
accordingly. Thishistory
is also known as
that analyze the data of ustomers
interest. Aftef that, thhey urate and shm
For Data target advertisirvg
Science, usually the data is
stored in different
Some of the formats. collected in the form of tables These tabular
1. CSV: commonly used formats are:
datasets can te
CSV
tabular data.stands
Each
for Comnma
line of this fileSeparated
fields which are is a data Values. t is a
comma, hence theyseparated
are
by
cormnas. record and eachsirnple
Since record
file forrnat used to store
2. known as CSV the values of onsists of one or rnore
Spreadsheet:
in
as form Spreadsheet
tabular
such
is a
files.
using rows and computer
records are separated by a
Microsoft colurmns. application used for anatysis and
SQL: Structured Excel, LibreOffice
3. Calc,There
etc.
are many
storage of daa
soreadshet rograms
used for storing, Query Language is one of availabie
System (RDBMS)manipulating and
retrieving
such as MySql, SOL
the nost
the data inpopular languages for
UNIT -III Server, etc. Relational Database DeMS. It is
1.1 INTRODUCTION BASIC ICT SKILLS Maragement
An
Operating system is a
operating system is its program
(EMPLOYABI LITY SKILLS)
automatically the usermost
designed
important
to run other
1.2
when switch prograns on a computer. A
on the program Operating
FUNCTIONS OF THE computer. prograns on a cornputer'
system s
1. It
manages all the
OPERATING SYSTEM starts
whether it is busy or devices
not.
of a
computer and keeps track of the
2. It also
3. It also
checks whether the device is status of the device,
controls software functioning
resources the conputer.
of property or not.
ASS X
use by
4. It manages the computer memory and keeps track of which memory spaceis in
which program and which space is free.
5. It keeps track of the amount of disk space used by a specific file
6. It allows you to create, copy, move and delete files.
1.3TYPES OF OPERATING SYSTEMS
Broadly there are two types of operating system interfaces:
1. Command Line Interface (CLI)

2. Interactive Operating System Interfaces


1.3.1 Command Line Interface

system where users type in commands using


A Command line interface is an older style operating
screen where user can type the desired
keyboard. Computer normally displays a prompt on a
command.

1.3.2 Interactive Operating System Interfaces


by
operating systems allow users to give commands without typing. ie.. just
The interactive sub-categories of interactive operating systems.
pointing or touching. There are two
(WIMP Interfaces)
(a) Graphical User Interface (GUI) user can use a
interfaces in which instead of typing in commands, the Windows, lcons,
GUI'sare visual (graphical) screen. The main features of a GUI
are
objects on the
mouse to point and click Operating Systems interactive in nature.
these
Menus and Pointers, which make
[Post-WIMP Interfaces]
(b) Touchscreen Interface screen.
interactive in nature and provide access via touch on
also touch
These operating systems are Interface include many finger based gestures apart from
Key features of a Touchscreen Rotating of screen, swiping on screen etc.
which are: Pinching of screen,
OPERATING SYSTEM
1.4 COMMONLY USED
popular computer operating system. It is the most famous
of a
1. Linux: Linux is the name open source development.
example of free software and
name of several families of GUI (Graphical User
Windows is the an operating
2. Windows: Microsoft Microsoft. (Microsoft first introduced
Interface) operating systems byNovember 1985 in response to the growing interest in
environment named Windows in
graphical user intefaces (GUls).
Linux: BOSS (Bharat Operating System
Operating System Solutions)
3. BOSS (Bharat distribution of GNU/Linux based on the
Debian.
Solutions) Linux is an Indian

[7)
ASS
X

4. Solarle: The Solaris Operating Systom, uBually known simply as Solaris, is a tree UnA
CLASS X

based operating systom intoduced by Sun Microsvstems


in 1992 as the
SunOS.

Mobilo Oporating Systom8: Those days mobile devices have turned smart, Gourtes,
6. operating systems are: Andro
smart mobile operating systoms. Some COmmon mobile
Apple iOS eto.

and LINUX are examples of distributed operating systems.


Windows, UNIX,
ICONS, AND TASK BAR ON THE
MENU, DESKTOP
1.5
1.5.1 Windows 70S

Windows 7is an
operating system developed by Microsoft and is used on personal computers
of Windows 7 Dosktop
1.5.1.1 Compononts
The components of Windows 7 Desktop are:
1. My Computor/ This PC: It displays all the storage areas of the computer.
2. Taskbar: Taskbar is the long horizontal bar present at the bottom of the screen.

3. Start button: It
is located on the left of the taskbar. When we Click the Start button, the
Stat menu opens and proVides access to programs and features.

A Recvcle Bin: Files and tolders deleted by the user are stored in Recycle Bin.
FOLDERS
1.6 FILES AND
The directory is also known as Tolder. The folder/directory structure is basically a storage location
or sub folders.
where we save our files
storane
Each directorv or folder can have files and sub-directory or sub-folders. A file is a named
Iocation where data is stored. This way of organizing files or folders is known as hierarchical file
that identifies the fle vne
structure. Each file is given a file name and has a file name extension
1.7 IMPORTANCE OF CARE AND
MAINTENANCE OF COMPUTER

Computer is an electronic device so it needs maintenance for its effective functioning. If the
maintenance is poor than there are chances of system failure.
Computer virus, malware, spyware, Trojan and other harmful software can ham the computer
system and lead to failure of hardware. If antivirus is installed in the computer, and is updated and
run periodically, then we can save our data from any kind of loss.
1.8BASIC TIPS FOR TAKING CARE OF DEVICES
Computer components needs proper care so that the devices stay safe for a longer duration.

Some of the precautions to be taken while cleaning the computer components are:

(8)
Never spray or
the liquid onto a squirt any
cotton cloth.liquid onto any computer
component. If a spray is neede0. spra
When cleaning a
component of the computer, turn it off before
When cleaning, be careful to not cleaning.
accidentally adjust any knobs or controis
UNIT- GREEN SKILLS
V
1.1 SUSTAINABLE DEVELOPMENT (EMPLOYABILITY SKILLS)
SUstainable development is a development that meets the needs of the present without
Compromising the ability of future generations to meet their own needs.
our dally lite we can contribute to create a Sustainable Society by following 4Rs' and 1U of
Sustainability.
Inese are REFUSE, REDUCE, REUSE, RECYCLE and UPCYCLE.
1. Refuse: We must first REFUSE to use products that may harm the environment. This wil
help in creating sustainable environment.
2. Reduce: REDUCE comes after REFUSE. In this step, we minimize the use of the products
that may cause harm to environment.

3. Reuse: Next Step is REUSE. In reusing we can use that resources in another way and
keeps them out of the trash.
we must try to
4. Recycle: Next step after REUSE is RECYCLE. After reusing the product,
recycle it, so that the leftover resource is used properly again.
5. Upcycle: UPCYCLING of products could be done manually asdesirable.well as with the help of
machines, giving a new look to the old product and making it look
DEVELOPMENT
1.2 IMPORTANCE OF SUSTAINABLE
resources
1. To keep a check on the over exploitation of natural
2. To regenerate the renewable energy resources
3. To ensure safety for human beings
4. To stop environmental degradation
5. Tofind alternative sources of energy

6. To utilize the resources to the optimum level


1.3DIMENSIONS OF SUSTAINABLE DEVELOPMENT
Sustainability is meeting the needs of the present without compromising the ability of future
generations to meet their needs. It has three main pillars: economic, environmental, and social
These three pillars are informaly referred to as people,planet andprofits.
[9)
CLASS X
1. The Environmental Pillar: The Environmental pillar is one of the most
ail three. The environmental pillar refers to the laws, regulations,
important
and other pilars
policy solid armong
Wate pollution, nechanisms
COncerning environmental issues These issues include air and
protection of pat
management, ecosystem management. maintenance of biodiversity, and
resources,wildife and endangered species.
social system, such as a country
2. The Social Pillar: Social Sustainability is the ability of a well-being and harmony. Problems
family, or organization, to function at a defined level of social
are symptoms of a system
like war, endemic poverty, widespread injustice, and low education rate the workers or weaker section
development of
that is socially unsustainable. Social pillar involves
society.
Of the economy, welfare of the local communities of the
they
3. The Economic Pillar: The economic pillar of sustainability is where most businesses feel
Activities that fit under the
are on fim ground. To be sustainable, a business must be profitable.
economic pillar include proper governance and risk management. This pillar is also referred to as
the governance pillar, referring to good corporate governance.
1.6 MEASURES TAKEN FOR IMPLEMENTING SDGS IN INDIA
Many programmes were introduced by the government of India like Swachh Bharat, Make in india,
Digital India and Skill India for promoting the SDGs.
Some other measures taken by India are as follows:
1. The ministry of statistics and programme Implementation (MOSPI) has been made, whicn Is
engaged in the process of making national indicators for the SDGs.
2. State Governments have been asked to pay keen attention on visioning, planning,
budgeting and developing monitoring system for SDGs.
3. States have been directed to implement and ensure the action, implementation and
monitoring of centrally sponsored schemes for implementing SDGs.
1.7 MEASURES TO OVERCOME CHALLENGES
Various measures have been taken to overcome the challenges are:

1. Organization with proper leadership to be made responsible: Government should


clearly specify an organization which should look after the implementation of SDGs.
2. Society to be included as a part and parcel of development: Societies at the state level
should be included in the proper execution of the SDGs.
3. Scope should be defined properly: A wel-defined outline of the SDGs should be
introduced so that the policy approach and execution should be done in a clear manner.
4. Indicators should be clearly stated: Clear indicators should be mentioned in the policy
planning.
5. Collection of correct and huge data to be made: Progress can be monitored only when
the large data collected is correct.
[10]

You might also like