Class 16 - DATA SCIENCE PROCESSES Notes from the AI Advance course by Irfan Malik & Dr Sheraz Naseer (Xeven Solutions)

DATA SCIENCE PROCESSES

Notes from the AI Advance course-Class 16 by Irfan Malik & Dr Sheraz Naseer (Xeven Solutions)

Future of Data Science is very Promising.

Emerging Field means, it has Buyer's.

Check out market trends & analysis.

Focus & Hardwork are necessary for Success.

Today is the war of Data.

AI is applicable in every field of Life.

Now, every industry is using AI to become more efficient.

AI WILL NOT TAKE YOUR JOB, BUT THE PERSON USING AI WILL.

Data Science Processes:

1. Ask

2. Prepare

3. Process

4. Analyze

5. Act

6. Share

Distribution in Data:

Gaussian Distribution

Normal Distribution (Bell Curve), also called Symmetrical Distribution of data.

Positively Skewed, Negatively Skewed

By Calculating Mean, Median & Mode. We can find out whether the data is Symmetrical or Asymmetrical.

If Mode < Mean, then data is Positively Skewed.

If Mode > Mean, then data is Negatively Skewed.

Follow the world, Follow the money.

Check out, what the world is doing.

Assignment: Go out in the market, Take data sample, Check out the parameters. It will help you to connect locally.

Start Thinking in AI. Develop Critical Thinking.

When you ask businesses regarding Data, don't ask directly regarding data, ask about sample, tell them regarding their benefit, that's called Business Development.

2 Reasons of Failure:

Implementation

Not know how to sell

You have to start Practicing Now.

Practice, Critical Thinking, Selling (All of 3 play their role)

Data Cleaning:

Remove Inconsistencies

Find out missing values

Data Integration & Data Reduction

Data Transformation is also very important in Prepare Process:

1- Smoothing (Regression is used)

2- Feature Engineering (Chances to Build Companies in this)

3- Normalization

Update Yourself on daily basics.

Feature Engineering Constructing new meaningful features that helps models.

In Pakistan, we have alot of Systems that are running without AI, ML, DS etc.

Pattern Extraction (Thoughts Regarding Data)

Behavior Change, Alone Time, Observing Patterns etc.

It is Traning your Brain, that's why we called it Data Science.

Practice, Practice, Practice.

Models are like maths equations.

Normalization:

Technique of Data Transformation

z score normalization (Mean, Standard Deviation). Also called z scaling & standard scaling.

min-max normalization (decide the range of values)

decimal scaling (processing point values)

That's the road map of Data Science. Now, you have to search more & more.

Also, Check out the Basic Course on Irfan Malik YouTube Channel to get detail knowledge regarding DS.

Analyzing Data:

For numeric data we use Regression (Linear, Logistics, Random Forest, Decision Trees etc.)

To Predict or Generate Category (Classification is used like K-Nearest Neighbor, Decision Trees, Bagging & Bosting)

Select the Techniques, Build the Model, Select the Best Model (Model Validation)

Confusion matrix involves Accuracy, Precision, Recall, F1.

Data & analysis will drive decisions.

Training model is a cyclic process.

Theoretical Knowledge will help you in long run.

Just Learn to ASK Questions.

Success makes noise. Work Hard & Believe on ALLAH.

Data Reporting:

In the form of Vizualization, Story Telling with Graphs.

Summarize & Concise Viz

Tools:

Tableau

Power BI

You can also use Python libraries for Viz.

It is wrong Perception that you can only learn when you hire for Job or Internship.

Go to Job as a useful.

Create something tangible, well organized, etc.

CEO look for his own benefit.

You can start Freelancing after this course.

Strong Intention, Commitment is Necessary for Success. First Step is very Important for Success.

#AI #artificialintelligence #datascience #irfanmalik #drsheraz #xevensolutions #hamzanadeem