Easy ways to start and to know about Machine Learning ,
Let see example below to visualize how above steps in action ,
Suppose we want to determine difference between Wine and Beer. How do we differensiate in terms of Color or Alcohol percentage ? . We need to have a data from different glassess of Beer and Wine .
Prepare a data in table format or graph in terms of color , % of Alcohol and Beer or Wine. Need to a choose a model as there different models created by Data scientists for Text, Speech , Images , colors and sequences. Then it comes a Traning which is most important part of ML. By Choosing a model we are able to differensiate between Beer and wine rather than depends of human judgement.
For training we have equation
y (output) = m (slope) * x (input) + b (Y-intercept)
weights = [m1,1 m1,2]
biases = [b1,1 b1,2]
Model is nothing but random values of matrix [W,b]
Training data -> Model [W,b] -> Prediction -> Test &Update W,b = Repetitive process
Actual steps of ML
1. Gathering Data
2. Preparing that Data
3. Choosing a Model
6. Hyperparameter Tuning
Appplications of Machine learning
2. Self driving cards
3. Detecting escalators in need of repair
4. Diabetic Retinopathy
It will rech $8.81 Billion by 2022 as per reports published by MarketsandMarkets
Want to get your hands dirty in ML use this API ,
TesorFlow is an open source software library for Machine Intelligence. Here is to play around API.
Image source: webtrends