kDimensions

Bite-sized ideas on AI in simple visuals.

[kDimensions @kdimensions1]

"The book is like a CEO summary of deep learning." -- Ronald T. Kneusel, Ph.D. (author of "Practical Deep Learning: A Python-Based Introduction")

You can now read a free sample of A Visual Introduction to Deep Learning here: https://t.co/5W3cLI5MxH

#deeplearning #learningisfun

[kDimensions @kdimensions1]

Performance measures in machine learning.

-- Cost - internal (for the model)
-- Metric - external (for the modeler)

[kDimensions @kdimensions1]

Frequentists and Bayesians.

[kDimensions @kdimensions1]

"But it works on my machine!"

Containerization
-- packages up software code and all its dependencies so it can run consistently on any infrastructure.
-- allows developers to create and deploy applications faster and more securely.

Source: IBM [https://t.co/1CDrDNaMgw]

[kDimensions @kdimensions1]

Source: Evidently AI.
Read the full article: https://t.co/MNybkiFjyc

[kDimensions @kdimensions1]

Data Drift and Concept Drift in Machine Learning.

Machine learning model performance degrades over time. When data quality is fine, there are two usual suspects: Data Drift or Concept Drift.

[kDimensions @kdimensions1]

Concept Drift: In contrast to the data drift, the distributions might even remain the same. Instead, the relationships between the model inputs and outputs change. In essence, the very meaning of what we are trying to predict evolves.

[kDimensions @kdimensions1]

Culture eats strategy for breakfast. — Peter Drucker

[kDimensions @kdimensions1]

Benefits TinyML over Cloud ML:
-- Low Latency
-- Reduced Power Consumption
-- Improved Environmental Performance
-- Network Bandwidth Efficiency
-- Strong Privacy

Full article by Edge Impulse: https://t.co/76TI76f17q

[kDimensions @kdimensions1]

TinyML

Running ML in the cloud is not always the best option, especially for applications that require low-latency and must be run close to the end-users.

TinyML -- ML models that run on very small pieces of hardware like ultra-low power microcontrollers.

[kDimensions @kdimensions1]

Learning and doing.

[kDimensions @kdimensions1]

Difference between Descriptive and Inferential Statistics

-- Descriptive statistics - Describes a sample.
-- Inferential statistics - Takes data from a sample and makes inferences about the larger population from which the sample was drawn.

Source: https://t.co/kQeBcvPV9R

[kDimensions @kdimensions1]

A great article on Causal Inference by Hugo Bowne-Anderson and Mike Loukides.

"What Is Causal Inference? An Introduction for Data Scientists".

Link: https://t.co/96zNGGDMbi

[kDimensions @kdimensions1]

5 - What should I do next?

[kDimensions @kdimensions1]

As amazing as it is, there are only five questions it can answer.

Source: @_brohrer_
[https://t.co/lZzIz8RaGp]

[kDimensions @kdimensions1]

4 - How much – or – How many?