Introduction to Deep Learning

There are two ways of looking at deep learning and its relationship to the larger field of artificial intelligence.

Deep learning can be described as a subset if machine learning which uses artificial neural networks. Machine learning is the study and development of machine that can learn from data.

Now because not all deep learning and even not all machine learning is focused around the pursuing generalized artificial intelligence we sometimes find the relationship between deep learning and artificial intelligence described as shown in “grouping 1”.

Generalized artificial intelligence implies building sentient machines meaning make

The simplified way of looking at the relationship between deep learning and artificial intelligence is what see over here in “groupings 2”. Where deep learning is a subset of machine learning which is in turn a subset of artificial intelligence.

Supervised learning is a method of transforming one dataset into other dataset.

For example if we have a dataset of called  February weather and that recorded the weather patterns of the month February for last 1,000 year  another dataset called April weather that contains the weather pattern of the month april over the same time period,  a supervised learning algorithm might try to use one to predict the other.

If we are able to successfully train the supervised learning algorithm on the 1000 years of February and April Weather, we would be able to predict the weather of April in the future given the weather of February of the same year.

Unsupervised learning finds patterns within a dataset. Clustering a dataset into groups is a type of unsupervised learning.


transforms a sequence of datapoints into a sequence of cluster labels. If the unsupervised learing algorithm learns 3 clusters, it’s common for these labels to be the numbers 1–3. Each datapoint will be assigned to a number based on which cluster it is in. You can think of unsupervised learning as simply clustering.


Currently there are wo major libraries designed for optimized machine learning performance on arm cortex-m processors. The first one is cmsis-nn, cmsis stands for the cortex microcontroller software interface standard and the nn stands for neural network.

Cmsis-nn is a collection optimized neural network functions for cortex-m processors. The library is optimized because it takes full advantage of the arm cortex-m hardware machine use of the floating point unit and SIMD capabilities of the microcontroller. We shall treat CMISIS-NN thoroughly in its own section. CMSIS-NN is managed by ARM.

The other popular library is the cubeMX.AI is was is managed by STMicroelectronics. It also allows us to deploy optimized machine learning models on the ARM-Cortex-M processor however this one works with only microntrollers from STMicroelectronics. But of course with patience and determination is quite possible to make it work with any ARM-Cortex microcontroller.  

Add Comment

Your email address will not be published. Required fields are marked *

Phone: +447491265055
W12 7JN, London, United Kingdom
34A Frithville Gardens
Style switcher RESET
Body styles
Color settings
Link color
Menu color
User color
Background pattern
Background image