Scribd
Upload
9 views

Data Science Lecture 3 - 4

Uploaded by

Yashika
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
9 views

Data Science Lecture 3 - 4

Uploaded by

Yashika
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 8

Course: Data Science in IoT

(Course Code: 2220505)


Centre for Internet of Things

Program :
B.Tech. (Internet of Things)
Semester/ Year : Fifth
Sem/ Third Year
Session :
Classes By:
July- Dec’ 2024 Dr. Saurabh Kumar Rajput
Assistant Professor
1. IoT Analytics & Challenges

2
2. Challenges in IoT Analytics
1. Data Volume and Velocity Challenge: IoT devices generate massive amounts of data in real-time. Managing, storing,
and processing this data is a challange, especially when considering the scale of millions of devices.
2. Data Variety and Complexity: IoT data comes in various formats (structured, unstructured, semi-structured) from
different sensors, devices, and sources. Integrating and standardizing this data for analysis is complex.
3. Data Quality and Integrity: IoT devices might produce noisy, incomplete, or erroneous data due to sensor
malfunctions, connectivity issues, or environmental factors.
4. Security and Privacy Concerns: IoT networks are often vulnerable to cyberattacks, and the data generated may contain
sensitive information, raising concerns about privacy.
5. Latency and Real-Time Processing: Some IoT applications require real-time or near-real-time analytics, which can be
challenging due to network latency, data processing delays, or computational constraints.
6. Scalability: As the number of connected devices grows, the infrastructure and analytics solutions must scale
accordingly to handle the increased load.
7. Energy and Resource Constraints: IoT devices, especially those in remote or mobile settings, are often resource-
constrained in terms of power, storage, and processing capability.
8. Contextual Understanding: Interpreting IoT data often requires understanding the context in which the data was
generated (e.g., environmental conditions, user behavior), which can be difficult to capture and analyze.
9. Data Governance and Management: Managing the lifecycle of IoT data, from collection to deletion, while ensuring
data integrity and security is complex.
10. Skill Gaps and Expertise: IoT analytics requires expertise in various domains, including data science, cybersecurity,
hardware, and networking, which may be scarce.
3
3. Data Analytics Life Cycle

4
3. Data Analytics Life Cycle
Discovery/ Business understanding: What is the issue? What is the Problem that an
organization is facing?
Example: A company is not able to increase the sales in a particular duration of the year.
 Framing the problem.
 Deciding the resources that you have.
 Assigning and scheduling the task.
 Planning the process of data analytics

Data Preparation:
 Extraction/ transformation of raw data into acceptable format.
 Removing un-necessary information from the data.
 Data conditioning, data cleaning, data processing.

Model Planning:
 Deciding the best suitable model for data analysis.
 Clustering/ Regression/ Classification etc. .
5
3. Data Analytics Life Cycle
Model Building:
 Building and executing the model decided in the model planning.
 Model for best result.

Communication Results:
 Analysis of the results, whether they match with expected results or not?
 If the results do not match with expected results, go back to discovery phase.

Operationalize:
 Creating documentation.
 Document both the expected results and un-expected results both.

6
4. IoT Analytics for the Cloud

7
4. IoT Analytics for the Cloud

You might also like