The Machine Learning University Course Overview

The Machine Learning University is a comprehensive course that covers various aspects of machine learning, including supervised and unsupervised learning, neural networks, deep learning, computer vision, and text processing. The course consists of nine modules, each focusing on a specific topic.

Learning Approaches in Machine Learning

------------------------------------

The course begins by covering the learning approaches comprising supervised and unsupervised learning. It is essential to understand that class imbalance is a critical issue in machine learning, particularly when the number of samples belonging to different classes are not equal. This can lead to inherent bias in the model, resulting in poor performance. To tackle this problem, there are ways to address it, such as oversampling and undersampling.

Oversampling involves increasing the number of samples in the minority class, while undersampling reduces the number of samples in the majority class. For example, if a dataset has 100 samples in class A and 200 samples in class B, oversampling can increase the number of samples in class A to 200, making both classes have an equal number of samples.

Machine Learning Applications

---------------------------

The course then moves on to machine learning applications, covering various topics such as supervised and unsupervised learning. It is essential to understand that machine learning models need to be trained and evaluated to ensure they are performing well.

Computer Vision Applications

-------------------------

The next module focuses on computer vision applications, including image representation, neural networks, and components of neural networks. The course also covers the development of computer vision models, including underfitting and overfitting, as well as model evaluation techniques.

Convolution Filters and Neural Networks

--------------------------------------

A key component of neural networks is convolution filters, which are used to extract features from images. The course covers the basics of convolution filters, including padding, stride, and pooling.

Deep Learning

--------------

The next module focuses on deep learning, covering various topics such as recurrent neural networks, gated recurrent units, long short-term memory networks, single-headed attention, and multi-headed attention.

Recurrent Neural Networks and Long Short-Term Memory Networks

-----------------------------------------------------------

Recurrent neural networks are designed to handle sequential data, such as time series or text. The course covers the basics of recurrent neural networks and long short-term memory networks, which are a type of recurrent neural network that can learn long-term dependencies in data.

Transformers and Attention Mechanisms

--------------------------------------

The course also covers transformer-based models, including single-headed attention and multi-headed attention. These mechanisms enable the model to attend to different parts of the input sequence simultaneously.

Underfitting and Overfitting

---------------------------

One of the challenges in machine learning is underfitting, where the model is too simple to capture the underlying patterns in the data. The course covers techniques for avoiding underfitting, such as regularization and hyperparameter tuning.

Regularization and Hyperparameter Tuning

-----------------------------------------

Regularization techniques can help prevent overfitting by adding a penalty term to the loss function. The course covers various regularization techniques, including L1 and L2 regularization.

Hyperparameter Tuning is essential to find the optimal set of parameters for a machine learning model. It involves adjusting the hyperparameters to achieve the best performance on a validation set while avoiding overfitting.

Text Processing and Preprocessing

---------------------------------

The final module focuses on text processing and preprocessing, covering various techniques such as tokenization, stemming, and lemmatization.

Text Vectorization

------------------

Text vectorization is a critical step in text processing, where the input text is converted into a numerical representation that can be fed into a machine learning model. The course covers various techniques for text vectorization, including bag-of-words and word embeddings.

Conclusion

----------

The Machine Learning University course offers a comprehensive overview of machine learning, covering various topics from supervised and unsupervised learning to deep learning, computer vision, and text processing. With this course, learners can gain a deep understanding of the subject and develop practical skills in implementing machine learning models.

Subscription to the Machine Learning University Channel

-----------------------------------------------------

The article concludes by encouraging readers to subscribe to the Machine Learning University channel, which offers access to all nine modules of the course for free. The channel also provides updates on new releases and other machine learning-related content.

By following along with the course, learners can develop a solid foundation in machine learning and stay up-to-date with the latest developments in the field.

"WEBVTTKind: captionsLanguage: enthree two one welcome back to the data professor youtube channel my name is chanin nantan ahmad and i'm an associate professor of bioinformatics in this video i'm going to talk about free machine learning courses provided by amazon machine learning university and without further ado we're starting right now okay so yesterday i saw this on my tweet so it is from the amazon science it is saying that classes from amazon's machine learning university or mlu that was previously only available to employees are now available to the public via on-demand videos okay and so let's have a look at the web page so they're providing the general background about machine learning and its application so before we begin maybe a little bit about machine learning so machine learning can be thought of as like the brains of artificial intelligence so you might have heard that ai or artificial intelligence can do this do that intelligently so how can it really actually do it so it relies on the capabilities of machine learning algorithms to learn from the data and make decisions on the basis of such data and so we call this as data driven and so some of the practical utilization of machine learning by amazon includes the amazon go which is a supermarket where you can go in take the items that you need directly from the shelf and then there will be cameras installed throughout the store so it will record what you have taken and then you could just put it in your bag and then just walk out of the store so without even going through the cashier okay because everything is done automatically via computer vision and machine learning and so actually machine learning university was started by amazon in 2016 and so the objective of the machine learning university from amazon was to provide its staff members updated information about machine learning okay so let's have a look further and so what is the platform for the machine learning university by amazon and it is actually a youtube channel called machine learning university and so this is one of the first course on machine learning university and it is on natural language processing and the instructor of the course is sam sasera and he is a applied scientist at the machine learning university at aws or amazon web service deep engine science and the second course is machine learning university course on computer vision and the instructor is rachel who who is an applied scientist at the amazon web service deep engine science and the third course is on tabular data and the instructor of the course is paula grad dino who is a technical training specialist okay so i'm going to provide you the link of this machine learning university by amazon in the video description okay so let's have a look at the website let's click on machine learning university and we're going to be directed to this let's click on the main page okay so in the main page here you will see that the youtube page is called machine learning university and there is 4 240 subscribers and so this is going at a very exponential rate because considering that videos were just recently published so here we see that only a week ago videos were uploaded and so you're going to see that some of the videos the views are not yet high because it's only a week and there are several videos as you will see here so the video provided here are not uploaded in sequential order so you want to click on the playlist in order to have access to the three courses that i have mentioned and here you will see the tabular data playlist which is comprised of 16 videos and then you're gonna see that there is the natural language processing comprising of 26 videos and computer vision consisting of 27 videos so let's have a look at the first course tabular data and then you're going to see the 16 videos on the side panel here and the first lecture is an introduction the course introduction and it's only three minutes and then the second is going to be on introduction to machine learning okay so this will be very important for the newcomer to the field so you're going to get a glimpse of what machine learning has to offer and then you're going to learn about model evaluation so model evaluation is where you have already built your machine learning model and then you're going to have to evaluate the performance of your model okay like for example if you have regression if you have classification how are you going to evaluate whether your model is reliable they are robust and they have good performance okay so i notice here that the subsequent lectures are not really in order of difficulty so they're pretty random so you're going to see that they're going to teach you about model evaluation but they haven't really taught you how to build a model yet and then in an actual exploratory data analysis should come before model evaluation so i think it's just a matter of reorganizing reshuffling the order of the videos in the list okay so probably you might want to watch introduction to machine learning and then hop on to the exploratory data analysis and so eda will give you a general glimpse of the overall characteristics of the data so generally you will apply descriptive statistics and you'll also make use of data visualization to see the relative distribution of the data okay and then there will be k-nearest neighbors and then a lecture on looking ahead so i think this is probably kind of like the conclusion of this course so probably it should be at the bottom okay and then you have using jupyter notebook on sagemaker okay so safety maker is a platform by amazon web service for data scientists and so here 2.1 okay it's introduction okay so they have numbered it so i guess they're probably ordering it according to the numbers here 1.1 1.2 1.3 into 1.6 okay so probably they have their own reason for ordering it in such a way okay so let's have a look further all right and then they have section 2.1 and then they have the introduction so probably that was the first section and then this is the second section okay so probably they have already included information how to build the model in 1.2 and then they will talk about model evaluation okay so in section 2.1 they have an introduction again and then they're going to cover more advanced topics okay so here we have feature engineering and then we have three base models and then we have hyper parameter tuning and then a brief introduction about the amazon web service sagemaker okay and then they have section 3.1 an introduction and then a lecture on optimization regression models let's have a look what it says and regularization learning model optimization all right and then they have 3.3 is ensemble methods boosting and then finally neural networks and the popular term that we hear a lot these days is auto ml okay all right so this is the tabular data let's have a look at the playlist for other course let's have a look for the courses in natural language processing so there's 26 videos here so let's go over the videos that are provided here so 1.1 is the course introduction to natural language processing 1.2 is introduction to machine learning so probably you could take the course in whatever order that you like so it doesn't mean that you have to take tabular data first and then take natural language processing because they're providing the introduction directly in each of the courses here so they're pretty much independent of one another all right so 1.2 is the introduction to machine learning 1.1 was the course introduction and so 1.3 is machine learning applications 1.4 they're going to cover about the learning approaches comprising of the supervised and unsupervised learning okay so class imbalance is very important particularly when your number of samples that are belonging to class a and class b are not equal and when they're not equal then the machine learning algorithm might be a bit confused in classifying the data samples so there will be inherent bias by the model so in order to tackle the problem of class imbalance then there are ways for you to do that such as over sampling and under sampling so in oversampling you probably will have to up the number of samples in the class that have the fewer samples for example if you have 100 data samples in class a 200 data samples in class b then you want to increase the number of data samples in class a from 100 to become 200 so now they're both 200 okay and so in under sampling you're going to reduce the number of data samples in class b from 200 to 100 so now it has the same number at 100. okay so let's continue and then 1.6 will be on missing values 1.7 model evaluation 1.9 machine learning and text 1.10 text pre-processing 1.11 text vectorization 1.12 again is k nearest neighbors and then they're going to have using jupiter notebook on sagemaker so i think it's probably the same video as the previous one and then 2.1 is tree based model 2.2 is regression models 2.3 optimization 2.4 regularization 2.5 hyper parameter tuning 3.1 neural networks 3.2 is word vectors let's talk about hyperbola all right and then now we have deep learning 3.3 recurrent neural networks 3.4 gated recurrent units 3.5 we have long short term memory networks all right and then we have 3.6 which is transformers seven is single headed attention eight multi-headed attention okay so that's the last video in the natural language processing let's head on back let's have a look at the last one and it is on computer vision okay so there's 27 videos so welcome so here we have the introduction to the course then we have introduction to machine learning and then we have machine learning applications we have supervised and unsupervised learning we have also again imbalanced data so this is again similar to the other course where it is on class imbalance they're the same thing and then 1.6 will be on underfitting overfitting and model evaluation so in the development of your machine learning models you try to avoid underfitting and overfitting and so the best way is to have your model generalized well on your data here is computer vision applications image representation neuron and activation functions neural networks components and training so you're going to see that in all of the courses they're going to cover neural networks in pretty much different difficulty level from the beginner's level to the more advanced 1.11 convolution filter padding stride and pulling and then again using jupyter notebooks on sagemaker so they're probably as i mentioned it's the same video first coding our different playlist 2.1 is computer vision data sets 2.2 lynnette 2.3 alex then 2.4 transfer learning okay now it's 3.1 vgg and batch normalization 3.2 is resnet the 20th course 3.3 is object detection applications 3.4 bounding box and anchor box and then sliding window method and non-max suppression 3.6 region based convolutional neural network okay in 3.7 they're going to cover the yolo or you only look once model 3.8 is semantic segmentation 3.9 fully convolutional networks and the last video is on unit all right so there you have it you have currently three courses that are available for you to listen to or watch so msn mentioned somewhere in the article that three are currently provided and here nine more will come before the end of the year so feel free to subscribe to the machine learning university channel which i have also done and you're going to have access to the remaining nine courses that are going to be released by the end of the year so the great thing about this is that it's free and you could follow along when you have spare time and so please make some spare time to learn machine learning if you're finding value in this video please smash the like button subscribe to the channel and hit on the notification bell in order to be notified of the next video and as always the best way to learn data science is to do data science and please enjoy the journey thank you for watching please like subscribe and share and i'll see you in the next one but in the meantime please check out these videosthree two one welcome back to the data professor youtube channel my name is chanin nantan ahmad and i'm an associate professor of bioinformatics in this video i'm going to talk about free machine learning courses provided by amazon machine learning university and without further ado we're starting right now okay so yesterday i saw this on my tweet so it is from the amazon science it is saying that classes from amazon's machine learning university or mlu that was previously only available to employees are now available to the public via on-demand videos okay and so let's have a look at the web page so they're providing the general background about machine learning and its application so before we begin maybe a little bit about machine learning so machine learning can be thought of as like the brains of artificial intelligence so you might have heard that ai or artificial intelligence can do this do that intelligently so how can it really actually do it so it relies on the capabilities of machine learning algorithms to learn from the data and make decisions on the basis of such data and so we call this as data driven and so some of the practical utilization of machine learning by amazon includes the amazon go which is a supermarket where you can go in take the items that you need directly from the shelf and then there will be cameras installed throughout the store so it will record what you have taken and then you could just put it in your bag and then just walk out of the store so without even going through the cashier okay because everything is done automatically via computer vision and machine learning and so actually machine learning university was started by amazon in 2016 and so the objective of the machine learning university from amazon was to provide its staff members updated information about machine learning okay so let's have a look further and so what is the platform for the machine learning university by amazon and it is actually a youtube channel called machine learning university and so this is one of the first course on machine learning university and it is on natural language processing and the instructor of the course is sam sasera and he is a applied scientist at the machine learning university at aws or amazon web service deep engine science and the second course is machine learning university course on computer vision and the instructor is rachel who who is an applied scientist at the amazon web service deep engine science and the third course is on tabular data and the instructor of the course is paula grad dino who is a technical training specialist okay so i'm going to provide you the link of this machine learning university by amazon in the video description okay so let's have a look at the website let's click on machine learning university and we're going to be directed to this let's click on the main page okay so in the main page here you will see that the youtube page is called machine learning university and there is 4 240 subscribers and so this is going at a very exponential rate because considering that videos were just recently published so here we see that only a week ago videos were uploaded and so you're going to see that some of the videos the views are not yet high because it's only a week and there are several videos as you will see here so the video provided here are not uploaded in sequential order so you want to click on the playlist in order to have access to the three courses that i have mentioned and here you will see the tabular data playlist which is comprised of 16 videos and then you're gonna see that there is the natural language processing comprising of 26 videos and computer vision consisting of 27 videos so let's have a look at the first course tabular data and then you're going to see the 16 videos on the side panel here and the first lecture is an introduction the course introduction and it's only three minutes and then the second is going to be on introduction to machine learning okay so this will be very important for the newcomer to the field so you're going to get a glimpse of what machine learning has to offer and then you're going to learn about model evaluation so model evaluation is where you have already built your machine learning model and then you're going to have to evaluate the performance of your model okay like for example if you have regression if you have classification how are you going to evaluate whether your model is reliable they are robust and they have good performance okay so i notice here that the subsequent lectures are not really in order of difficulty so they're pretty random so you're going to see that they're going to teach you about model evaluation but they haven't really taught you how to build a model yet and then in an actual exploratory data analysis should come before model evaluation so i think it's just a matter of reorganizing reshuffling the order of the videos in the list okay so probably you might want to watch introduction to machine learning and then hop on to the exploratory data analysis and so eda will give you a general glimpse of the overall characteristics of the data so generally you will apply descriptive statistics and you'll also make use of data visualization to see the relative distribution of the data okay and then there will be k-nearest neighbors and then a lecture on looking ahead so i think this is probably kind of like the conclusion of this course so probably it should be at the bottom okay and then you have using jupyter notebook on sagemaker okay so safety maker is a platform by amazon web service for data scientists and so here 2.1 okay it's introduction okay so they have numbered it so i guess they're probably ordering it according to the numbers here 1.1 1.2 1.3 into 1.6 okay so probably they have their own reason for ordering it in such a way okay so let's have a look further all right and then they have section 2.1 and then they have the introduction so probably that was the first section and then this is the second section okay so probably they have already included information how to build the model in 1.2 and then they will talk about model evaluation okay so in section 2.1 they have an introduction again and then they're going to cover more advanced topics okay so here we have feature engineering and then we have three base models and then we have hyper parameter tuning and then a brief introduction about the amazon web service sagemaker okay and then they have section 3.1 an introduction and then a lecture on optimization regression models let's have a look what it says and regularization learning model optimization all right and then they have 3.3 is ensemble methods boosting and then finally neural networks and the popular term that we hear a lot these days is auto ml okay all right so this is the tabular data let's have a look at the playlist for other course let's have a look for the courses in natural language processing so there's 26 videos here so let's go over the videos that are provided here so 1.1 is the course introduction to natural language processing 1.2 is introduction to machine learning so probably you could take the course in whatever order that you like so it doesn't mean that you have to take tabular data first and then take natural language processing because they're providing the introduction directly in each of the courses here so they're pretty much independent of one another all right so 1.2 is the introduction to machine learning 1.1 was the course introduction and so 1.3 is machine learning applications 1.4 they're going to cover about the learning approaches comprising of the supervised and unsupervised learning okay so class imbalance is very important particularly when your number of samples that are belonging to class a and class b are not equal and when they're not equal then the machine learning algorithm might be a bit confused in classifying the data samples so there will be inherent bias by the model so in order to tackle the problem of class imbalance then there are ways for you to do that such as over sampling and under sampling so in oversampling you probably will have to up the number of samples in the class that have the fewer samples for example if you have 100 data samples in class a 200 data samples in class b then you want to increase the number of data samples in class a from 100 to become 200 so now they're both 200 okay and so in under sampling you're going to reduce the number of data samples in class b from 200 to 100 so now it has the same number at 100. okay so let's continue and then 1.6 will be on missing values 1.7 model evaluation 1.9 machine learning and text 1.10 text pre-processing 1.11 text vectorization 1.12 again is k nearest neighbors and then they're going to have using jupiter notebook on sagemaker so i think it's probably the same video as the previous one and then 2.1 is tree based model 2.2 is regression models 2.3 optimization 2.4 regularization 2.5 hyper parameter tuning 3.1 neural networks 3.2 is word vectors let's talk about hyperbola all right and then now we have deep learning 3.3 recurrent neural networks 3.4 gated recurrent units 3.5 we have long short term memory networks all right and then we have 3.6 which is transformers seven is single headed attention eight multi-headed attention okay so that's the last video in the natural language processing let's head on back let's have a look at the last one and it is on computer vision okay so there's 27 videos so welcome so here we have the introduction to the course then we have introduction to machine learning and then we have machine learning applications we have supervised and unsupervised learning we have also again imbalanced data so this is again similar to the other course where it is on class imbalance they're the same thing and then 1.6 will be on underfitting overfitting and model evaluation so in the development of your machine learning models you try to avoid underfitting and overfitting and so the best way is to have your model generalized well on your data here is computer vision applications image representation neuron and activation functions neural networks components and training so you're going to see that in all of the courses they're going to cover neural networks in pretty much different difficulty level from the beginner's level to the more advanced 1.11 convolution filter padding stride and pulling and then again using jupyter notebooks on sagemaker so they're probably as i mentioned it's the same video first coding our different playlist 2.1 is computer vision data sets 2.2 lynnette 2.3 alex then 2.4 transfer learning okay now it's 3.1 vgg and batch normalization 3.2 is resnet the 20th course 3.3 is object detection applications 3.4 bounding box and anchor box and then sliding window method and non-max suppression 3.6 region based convolutional neural network okay in 3.7 they're going to cover the yolo or you only look once model 3.8 is semantic segmentation 3.9 fully convolutional networks and the last video is on unit all right so there you have it you have currently three courses that are available for you to listen to or watch so msn mentioned somewhere in the article that three are currently provided and here nine more will come before the end of the year so feel free to subscribe to the machine learning university channel which i have also done and you're going to have access to the remaining nine courses that are going to be released by the end of the year so the great thing about this is that it's free and you could follow along when you have spare time and so please make some spare time to learn machine learning if you're finding value in this video please smash the like button subscribe to the channel and hit on the notification bell in order to be notified of the next video and as always the best way to learn data science is to do data science and please enjoy the journey thank you for watching please like subscribe and share and i'll see you in the next one but in the meantime please check out these videos\n"