Pathway to ML:
In real world we have data problems that we need to solve
Or we have data questions that we have to answer
Here in the pathway, we have real world.
- From where we get the raw data it can be from anything like sensors, survey etc.
- Then we just store that data into some files like csv, cloud etc.
The above two points terms as the Data Engineering (collecting and storing the data)
- After collecting the data we just have to clean it like finding the missing values, recognising the data, restructure data etc.
- Then we have Exploratory data analysis which includes data visualization
The above two points terms as the Data Analysis or Scientist (cleaning and visualizing the data)
- After the analysis we went towards the ML models
- Supervised Learning: predict on outcome
- Unsupervised Learning: discover patterns in data
The above 3 points are together consider as Data Scientist or Machine Learning Engineer
- After all the above steps we create a data product which include (service, dashboard, application) by this data product we use to predict the future outcomes and we can gain insights on the data
NumPy:
All the Data Science libraries like skit learn, seaborn, etc. are built using NumPy
It is an N-dimensional Array
It looks similar to the python list but it has much more efficiency
Broadcasting capabilities are useful for quickly applying functions to the datasets
Creation: