I’ve been programming for almost 10 years.
This probably means I don’t take courses anymore, right?
Wrong, I still do.
Recently, I failed a machine learning course. But I still learned a lot - ML is extremely powerful. Machine Learning involves making computers improve on a selected task by learning from data, instead of explicitly coding the rules.
There are many kinds of machine learning systems, some of the main ones are:
Machine Learning projects involve gathering data in a training set, feeding the training set to the learning algorithm, then depending on the type, the algorithm learns from the training set. Model-based algorithms modify the model to the training set. Instance-based algorithms use similarity measures with the training set.
Training sets are crucial and the system will not perform well under conditions like the training set being too small, not representing the data correctly, or having too many bad examples in the training set (some noise is good and needed though!).
Finally, you must fine-tune and evaluate the model until you get the best results. This process can be tedious, but the slightest adjustment can make a very big difference.
Now while some people might think their road ends when they fail, my journey into machine learning is just beginning. Follow me to stay up to date on my learnings!