that score
Ads – scroll to continue
for next
action
assign
Ads – scroll to continue
Score
to yourself
action
Ads – scroll to continue
To evaluate
result
of that action
Ads – scroll to continue
based on
that state and
the environment
Ads – scroll to continue
…after that
I will do it
action
this machine
make
decision…
Five
Four
3
2
1
data in the same category
then you can process,
treated in a similar way
the model receives the data
without prior classification
Examples: video games, root
optimization, chatbot training,
self-driving car
e.g. customer segmentation,
recommendation system,
image compression
try to group them
by similarity
produce the most
Homogeneous category
Examples: anti-spam filters, disease
detection, image recognition
“Cat”
When new data is provided,
AI recognizes patterns
of known categories
The model identifies
Data commonalities
in the same category
dog
the model is being trained
About classified data
by humans
Source: Microsoft, IBM, Christina Bernard, Nvidia, Ai.nl
3 types of machine learning algorithms
The model is trained based on
unique performance and
and the previous result
reinforcement learning
unsupervised learning
