Machine Learning is not something that can be mastered or learnt in a short time, and you need at least 3-4 months familiarize yourself with the basics and even after that you need at least 6-7 months to get to a good place with your ML knowledge.
To get started you can first go through this course (Python for
Everybody - Full University Python Course) to ...
He can use no-code ML platforms such as: RapidMiner Studio, Google ML Kit, Orange, and BigML.
Also, this article is very good article for learning RapidMiner
The tf.data.Dataset.cache transformation can cache a dataset, either in memory or on local storage. This will save some operations (like file opening and data reading) from being executed during each epoch. The next epochs will reuse the data cached by the cache transformation.
Prefetch overlaps the preprocessing and model execution of a training step. While ...
You could try using the flow_from_directory() method on your ImageDataGenerator class, which does the augmentation - only a small change is necessary:
H = model.fit(
aug.flow_from_directory(trainX, trainY, batch_size=BS),
If you start using a tf.data.Dataset directly, you will get more control over how the data is read from disk (caching, ...
The current approach use 70/30 or 80/20, the most used is 80/20 (train/test). However there is other things you should check, for example if you data is balanced. If your data is not balanced you might want to use undersample or oversample.
There are many differences as these are inherently complete different products with different goals.
[+] cloud deployment comes out of the box (including a rest API)
[+] labelling tool to add data and label them
[-] you have no control over the learning algorithm
[-] difficult to run your model locally/completely for free
tensorflow (or any ...
From a practical point view, yes you could use both options seamlesly with in general similar results (check this scikit learn functionality), but:
SGD (Stochastic Gradient Descent) is an optimization algorithm among others
log loss/hinge loss... are the loss functions used in the selected optimization strategy (SGD in your case) to find the optimal weights ...
If you only consider the process of making something from your data (excluding data extraction etc), I'd say something like :
Visualising your data to understand it
Cleaning your data, removing useless features etc
Feature engineering : keeping only relevant features, creating new features from existing ones (combining a few one, making ratio of 2 features ...