We Have to Explore the Features: Analyze the Features in Order to Determine Which Features Were Important and With That Knowledge We Could Go Back and Collect Additional Data and Then Build Further Models in an Iterative Manner

Okay, so let's have a look further. In Lesson One, Problem Definition, it provides a brief overview of what the lesson one will be about. And then you're gonna see that the video is embedded right inside here. So you could just click on it. Let's have a look further. Lesson Two is on data, and as you know, data is the essential part of the data science process. And so in the video here, it will be talking about how to prepare the training data, and the video is embedded here, and so lesson three will be about model evaluation. So this will allow us to determine how robust how reliable how well did the model performed, and so this typically will involve looking at the model performance metrics, such as in a regression model, you will look at the rmse, the mse, which are essentially the error, and then also the r square.

In a classification model, you will look at accuracy, sensitivity, specificity, Matthew's correlation coefficient, area under the curve of the roc curve. Okay, so let's have a look further, and you see that the video is embedded right here. So let's try clicking on it. Okay, so you're gonna see that videos are taught by members of the ads team of Facebook, so the ads machine learning team of Facebook. So let's have a look at video two. Okay, so also software engineer of the apps machine learning team and the introductory video from the engineering manager of the ads machine learning team on Facebook.

Let's Take a Look Further in Lesson Four

Lesson four will take a deep dive into the features. So in a typical machine learning model, features are essentially descriptive variables that describe the quantitative or qualitative features of data samples that we are interested in. For example, if we're analyzing a data set of iris flowers, then the features would be the dimensions of the sepo and the petal part of the flower. If we're analyzing a data set of molecules where we want to predict the likelihood of a drug molecule becoming active or inactive, then the features will be the atomic constituent of the molecule, which will essentially be the molecular properties or the functional group of the molecule.

Or if we're analyzing health data, this could include the blood glucose level, the lipid profile of the blood, the diastolic systolic blood pressure, etc. So in a nutshell, features provide a means to describe the data samples of interest into numerical or qualitative terms, and then such data will be used by the machine learning algorithm in its learning process. And so feature engineering is an essential part of the machine learning model building process. So in lesson five, we will be about model selection. So this will include how to pick a model, how to tune a model, and how to compare models.

So for example, if you're building several machine learning models using different algorithms, maybe you're comparing between linear regression support vector machine random forest, and so how will you pick which model will be the most suitable. Or even before picking the model, how do you actually tune the model by tuning the model meaning that you're optimizing the parameters, the hyperparameters of the model. And also, you're not using only one machine learning algorithm, you're using several, and so you're comparing the performance of several machine learning algorithms.

And so the question is, how do you pick which model built by different machine learning algorithms are the best for use in a practical setting? So this video will cover that topic and issues. Okay, so before further continuing, let's click on the video and see briefly.

Six Videos from the Ads Machine Learning Team on Facebook

In Lesson Four, it is taught by a research scientist of the ads machine learning team on Facebook. And in five, taught by a software engineer of the apps machine learning team on Facebook. Okay, so finally in lesson six, it will be focusing on experimentation. So it is mentioned here that one of the key topic is the difference between offline and online experimentation.

Okay, so there you have it, six videos from the ads machine learning team on Facebook, and so each of these videos are about six to ten minutes in duration, and so in all you can finish all of the videos in about one hour. So that should give you a rough understanding of how the ads machine learning team of Facebook are using machine learning to get insights from the Facebook ads, and so it is a great opportunity to learn the fundamentals of the end-to-end project of the ads team and gain insights on the tips and tricks of how they are doing it.

And so if you're finding value in this video, please give it a thumbs up subscribe if you haven't yet done so 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 practice.

"WEBVTTKind: captionsLanguage: enthree two one welcome back to the data professor youtube channel my name is and i'm an associate professor of bioinformatics do you want to learn machine learning and do you want to learn it for free if you answered yes then you want to watch this video to the end because today i'm going to show you how you can learn machine learning directly from facebook and without further ado we're starting right now so in this series i share with you learning resources for data science and machine learning and today we're going to look at a machine learning resource from facebook and it is called the facebook field guide to machine learning video series so this series is comprised of a total of six videos and it encompasses all of the essential part of the data science process including setting the goals gathering the data preparing the data creating the models testing different parameters in the models testing various scenarios of the models and then evaluating the performance of the models okay and all of this are weaved together in what we call an experimentation so it's kind of like doing research when you're modifying some parameters of your interest like for example in a a b testing and then you're comparing the various models to see which one provided the best performance and which model provided the results that you are looking for so that result could be a performance increase a better drug which are the end outcome but it can also provide you insights into the underlying features of the models such as which features were important and this will provide us data-driven insights so that we can direct our actions toward the important features in hope of obtaining the desirable outcome okay so let's have a look at the topics of the video series okay so the link to this facebook field guide to machine learning will be provided in the video description and also in the pin comment or also in the card up above so let's take a look here so this paragraph describes about the importance of machine learning and artificial intelligence and this tutorial series by facebook was developed by the ads machine learning team of facebook and it explores six topic as i have mentioned and this includes problem definition data evaluation features model and experimentation so in a typical data science life cycle this starts with defining the problem that we want to explore and then we're going to collect the relevant data and then we're going to pre-process the data prepare the data or wrangle the data handle missing data split the data and then using subsets of the data to build a prediction model whether it could be a regression model a classification model and then finally evaluating the performance of the model where we could try different parameter settings different feature sets and finally in order to gain insights from the model we have to explore the features analyze the features in order to determine which features were important and with that knowledge we could go back and collect additional data and then build further models in an iterative manner okay so let's have a look further in the article here so you will see here that in lesson one problem definition it provides a brief overview of what the lesson one will be about and then you're gonna see that the video is embedded right inside here so you could just click on it let's have a look further so lesson two is on data and as you know data is the essential part of the data science process and so in the video here it will be talking about how to prepare the training data and the video is embedded here and so lesson three will be about model evaluation so this will allow us to determine how robust how reliable how well did the model performed and so this typically will involve looking at the model performance metrics so for example in a regression model you will look at the rmse the mse which are essentially the error and then also the r square in a classification model you will look at accuracy sensitivity specificity matthew's correlation coefficient area under the curve of the roc curve okay so let's have a look further and you see that the video is embedded right here so let's try clicking on it okay so you're gonna see that videos are taught by members of the ads team of facebook so the ads machine learning team of facebook so let's have a look at video two okay so also software engineer of the apps machine learning team and the introductory video from the engineering manager of the ads machine learning team on facebook okay so let's take a look further in lesson four so lesson four will take a deep dive into the features so in a typical machine learning model features are essentially descriptive variables that describe the quantitative or qualitative features of data samples that we are interested in so for example if we're analyzing a data set of iris flowers then the features would be the dimensions of the sipo and the petal part of the flower if we're analyzing a data set of molecules where we want to predict the likelihood of a drug molecule becoming active or inactive then the features will be the atomic constituent of the molecule which will essentially be the molecular properties or the functional group of the molecule or if we're analyzing health data then this could include the blood glucose level the lipid profile of the blood the diastolic systolic blood pressure etc so in a nutshell features provide a means to describe the data samples of interest into numerical or qualitative terms and then such data will be used by the machine learning algorithm in its learning process and so feature engineering is an essential part of the machine learning model building process and so in lesson five will be about model so this will include how to pick a model how to tune a model and how to compare models so for example if you're building several machine learning models using different algorithms maybe you're comparing between linear regression support vector machine random forest and so how will you pick which model will be the most suitable or even before picking the model how do you actually tune the model by tuning the model meaning that you're optimizing the parameters the hyperparameters of the model and also you're not using only one machine learning algorithms you're using several and so you're comparing the performance of several machine learning algorithms and so the question is how do you pick which model built by different machine learning algorithms are the best for use in a practical setting and so this video will cover that topic and issues so before further continuing let's click on the video and see briefly so in lesson four it is taught by a research scientist of the ads machine learning team on facebook and in five taught by a software engineer of the apps machine learning team on facebook okay and so finally in lesson six it will be focusing on experimentation and so it is mentioned here that one of the key topic is the difference between offline and online experimentation okay and there you have it six videos from the ads machine learning team on facebook and so each of these videos are about six to ten minutes in duration and so in all you can finish all of the videos in about one hour so that should give you a rough understanding of how the abs machine learning team of facebook are using machine learning to get insights from the facebook ads and so it is a great opportunity to learn the fundamentals of the end-to-end project of the abs team and gain insights on the tips and tricks of how they are doing it and so if you're finding value in this video please give it a thumbs up subscribe if you haven't yet done so 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 and i'm an associate professor of bioinformatics do you want to learn machine learning and do you want to learn it for free if you answered yes then you want to watch this video to the end because today i'm going to show you how you can learn machine learning directly from facebook and without further ado we're starting right now so in this series i share with you learning resources for data science and machine learning and today we're going to look at a machine learning resource from facebook and it is called the facebook field guide to machine learning video series so this series is comprised of a total of six videos and it encompasses all of the essential part of the data science process including setting the goals gathering the data preparing the data creating the models testing different parameters in the models testing various scenarios of the models and then evaluating the performance of the models okay and all of this are weaved together in what we call an experimentation so it's kind of like doing research when you're modifying some parameters of your interest like for example in a a b testing and then you're comparing the various models to see which one provided the best performance and which model provided the results that you are looking for so that result could be a performance increase a better drug which are the end outcome but it can also provide you insights into the underlying features of the models such as which features were important and this will provide us data-driven insights so that we can direct our actions toward the important features in hope of obtaining the desirable outcome okay so let's have a look at the topics of the video series okay so the link to this facebook field guide to machine learning will be provided in the video description and also in the pin comment or also in the card up above so let's take a look here so this paragraph describes about the importance of machine learning and artificial intelligence and this tutorial series by facebook was developed by the ads machine learning team of facebook and it explores six topic as i have mentioned and this includes problem definition data evaluation features model and experimentation so in a typical data science life cycle this starts with defining the problem that we want to explore and then we're going to collect the relevant data and then we're going to pre-process the data prepare the data or wrangle the data handle missing data split the data and then using subsets of the data to build a prediction model whether it could be a regression model a classification model and then finally evaluating the performance of the model where we could try different parameter settings different feature sets and finally in order to gain insights from the model we have to explore the features analyze the features in order to determine which features were important and with that knowledge we could go back and collect additional data and then build further models in an iterative manner okay so let's have a look further in the article here so you will see here that in lesson one problem definition it provides a brief overview of what the lesson one will be about and then you're gonna see that the video is embedded right inside here so you could just click on it let's have a look further so lesson two is on data and as you know data is the essential part of the data science process and so in the video here it will be talking about how to prepare the training data and the video is embedded here and so lesson three will be about model evaluation so this will allow us to determine how robust how reliable how well did the model performed and so this typically will involve looking at the model performance metrics so for example in a regression model you will look at the rmse the mse which are essentially the error and then also the r square in a classification model you will look at accuracy sensitivity specificity matthew's correlation coefficient area under the curve of the roc curve okay so let's have a look further and you see that the video is embedded right here so let's try clicking on it okay so you're gonna see that videos are taught by members of the ads team of facebook so the ads machine learning team of facebook so let's have a look at video two okay so also software engineer of the apps machine learning team and the introductory video from the engineering manager of the ads machine learning team on facebook okay so let's take a look further in lesson four so lesson four will take a deep dive into the features so in a typical machine learning model features are essentially descriptive variables that describe the quantitative or qualitative features of data samples that we are interested in so for example if we're analyzing a data set of iris flowers then the features would be the dimensions of the sipo and the petal part of the flower if we're analyzing a data set of molecules where we want to predict the likelihood of a drug molecule becoming active or inactive then the features will be the atomic constituent of the molecule which will essentially be the molecular properties or the functional group of the molecule or if we're analyzing health data then this could include the blood glucose level the lipid profile of the blood the diastolic systolic blood pressure etc so in a nutshell features provide a means to describe the data samples of interest into numerical or qualitative terms and then such data will be used by the machine learning algorithm in its learning process and so feature engineering is an essential part of the machine learning model building process and so in lesson five will be about model so this will include how to pick a model how to tune a model and how to compare models so for example if you're building several machine learning models using different algorithms maybe you're comparing between linear regression support vector machine random forest and so how will you pick which model will be the most suitable or even before picking the model how do you actually tune the model by tuning the model meaning that you're optimizing the parameters the hyperparameters of the model and also you're not using only one machine learning algorithms you're using several and so you're comparing the performance of several machine learning algorithms and so the question is how do you pick which model built by different machine learning algorithms are the best for use in a practical setting and so this video will cover that topic and issues so before further continuing let's click on the video and see briefly so in lesson four it is taught by a research scientist of the ads machine learning team on facebook and in five taught by a software engineer of the apps machine learning team on facebook okay and so finally in lesson six it will be focusing on experimentation and so it is mentioned here that one of the key topic is the difference between offline and online experimentation okay and there you have it six videos from the ads machine learning team on facebook and so each of these videos are about six to ten minutes in duration and so in all you can finish all of the videos in about one hour so that should give you a rough understanding of how the abs machine learning team of facebook are using machine learning to get insights from the facebook ads and so it is a great opportunity to learn the fundamentals of the end-to-end project of the abs team and gain insights on the tips and tricks of how they are doing it and so if you're finding value in this video please give it a thumbs up subscribe if you haven't yet done so 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"