Deep Learning playlist overview & Machine Learning intro

Key Takeaways

[Music] in this playlist will be covering multiple topics and deep learning will discuss different terms understand what they mean and how they fit into the overall deep learning framework and some videos will also discuss how certain topics are implemented in code I'll be using the Charis neural network API written in Python for these demos that show the technical implementation of the topics that we'll discuss I have a separate playlist called deep learning with Kerris on my channel that covers basics for getting started with Kerris many videos in that playlist assume that you already have a basic understanding of the deep learning topics that we'll be discussing here but if you want to follow along with the code implementations that we'll do in some of these videos then at least check out the Charis prerequisite video in the Charis playlist so you can get everything you need in order to make use of Karass going forward so in this intro video I'd like to start with the very basics like I said we'll be discussing deep learning so we're going to find out what that even means but to understand deep learning we first have to understand machine learning machine learning is the practice of using algorithms to analyze data and then learn from that data and then make a determination or prediction about new data so ok this almost sounds similar to what you do with regular code right so you'd write an algorithm the machine would execute the algorithm on particular data that you give it and then it could later do the same task with new data that it's never seen before sure so how does machine learning contrast to just writing a logical algorithm to instruct a computer on how to do something well in the definition of machine learning I just gave the focus is on this learn from that data part with machine learning rather than manually writing code with a specific set of instructions to accomplish a particular task the machine is trained using large amounts of data and algorithms that give it the ability to learn how to perform the task without being explicitly told how to do so so this sounds like magic but we'll see in later videos how this learning is done so first let's look at a really basic example of how something could be done with machine learning versus how the same task could be done with traditional programming so I have an example listed here that states that the task is to analyze the sentiment of a popular media outlet and classify the sentiment as positive or negative so basically you would want an algorithm to be able to look through different articles published by this media outlet and then determine whether the article was a positive article or a negative article so that's what we would be doing with this sentiment analysis example so with a traditional programming approach you may give your algorithm a list of words that could be classified as negative or positive so the words like can't or won't or unfortunately or sorry things of that nature might be classified as negative where things like happily thankfully great amazing might be classified as positive so you could give a list of positive words and negative words to an algorithm and tell the algorithm to tally up how many times words in the negative category versus words in the positive category are used in a particular article and then at the end your algorithm may just look and see which list is longer the positive list or negative list and whichever one is longer it would classify that article as positive or negative based on which one has the longer list with the machine learning approach however you wouldn't have an algorithm with an explicit set of instructions to follow like we just discussed with the traditional approach instead the algorithm would analyze large amounts of given media data and then learn the features from that data and classify what a negative article looks like verses a positive article you as a person wouldn't be explicitly coding any particular words for the algorithm to recognize instead the algorithm would learn that certain words are good or positive and learn that certain words are negative and then later you could give it new articles that it's never seen before and it could make a prediction about whether those articles are positive or negative based on what it's learned from the previous data that it was trained on so now that we understand generally what machine learning is the next topic we want to cover is deep learning and deep learning is a tool or technique that can be used to implement machine learning will be covering deep learning in detail in our next video so stay tuned for that I hope you found this video helpful if you did please like this video subscribe suggest and comment and thanks for watching [Music]