The Next Live Session: Time Series Analysis and Forecasting

The next live session that we will be conducting will be on the 5th of January, which is the coming Sunday from 10:00 a.m. to 12:00 p.m. This time the topic is time series analysis and forecasting, and this session will be accessible to all of our registered students. The desktop app and Slack will be used for all of the chat during the live communication during the live session itself.

We have covered many related topics around time series analysis and forecasting in our course videos. For example, we have discussed some very simple stuff in our taxi demand prediction problem, which is one of our case studies. In this case study, we talk about simple concepts like weighted moving averages and exponentially weighted moving averages. We also look at how to pause a time series forecasting problem as a typical regression problem.

Additionally, we discuss about 40's Fourier transforms or Fourier series on how to detect the repeated Nisour the odd-odd or the or at what intervals is a is the time series repeating. We have used some of these concepts extensively in the case study of taxi demand prediction. We also discussed ILSTM (Long Short-Term Memory) models in lots of detail, which are almost the state of the art for all time series forecasting problems.

However, still some of our students are confused on how to apply these techniques to real-world problems because all these concepts are spread around the course they are not all discussed at one place because it's not possible to do it to understand ILSTM so we have to cover all of deep learning but to do simple weighted moving averages and simple concepts basic concepts are good enough right to understand Fourier transforms a little bit of mathematical maturity is required so these concepts are spread around the whole course as and when the need arise.

This is slightly confusing for our students because the little cart of where a real-world problem comes up they're little caught up on what technique to apply which technique is most relevant here and things like that. So, what we thought we'll do is we'll take some real-world cases and in each of these cases we'll discuss which techniques to apply how to apply them gracefully and I will give pointers to the course videos where the concept has been discussed in lot more detail.

There are some classical techniques like ARIMA (AutoRegressive Integrated Moving Average) that we have not discussed in the course because today state-of-the-art in forecasting is all about ILSTMs. But, what I'll try to do in this live session as time permits we'll try and cover a few classical techniques also so that your breadth of knowledge as well as your conceptual understanding of alternative techniques also improves.

In this two-hour session, we will try to take a bunch of problems we will try to solve them using techniques that we've already discussed in the course videos at the same time as and when there is something or as and when there is a context where we can introduce a new concept or a new idea or a new method or technique we will discuss that in detail again. We will try to finish all of this in two live sessions if we run out of time we'll do one more live session on this topic.

This live session can be beneficial to students who are both in the early sections of the course videos as well as in the later sections of the course videos. Course students who are in the earlier sections of the course videos may not be able to understand when I talk about ILSTMs etc but they get a much more deeper understanding and contextual understanding on where ILSTMs are actually useful in the bigger picture of things right.

But, we will be using again I'll be referring to very simple basic data analysis and basic concepts which even students were in the earlier sections of the course can benefit from certainly students who are in the later sections of the course will get a much better picture because they would have already learned about ILSTMs they would have done all the case studies in the course and things like that right so we will try to do an X as extensive as possible live session on time series analysis if you cannot do it in one will do it in two live sessions okay see you this Sunday hoping to have a great session with all of you.

"WEBVTTKind: captionsLanguage: enhi friends the next life session that we will be conducting will be on the 5th of Jan which is the coming Sunday from 10:00 a.m. to 12:00 in the noon and this time the topic is time series analysis and forecasting this will be accessible to all of our registered students why the desktop app and as usual will you slack for all of the chat during the or any live communication during the during the live session itself having said that we have covered lots of related topics around time series analysis and forecasting the course videos itself for example we have discussed some very simple stuff in our taxi in a taxi demand prediction problem which is which is one of our case studies where we talk about simple concepts like weighted moving averages or exponentially weighted moving averages we also look at how to pause a time series forecasting problem as a typical regression problem again we also discuss about 40's Fourier transforms or Fourier series on how to detect the repeated Nisour the odd-odd or the or at what intervals is a is the time series repeating we have used some of these concepts extensively in the case study of taxi demand prediction also we also have discussed illicit iums in lots of detail which are almost the state of the art for all of timeseriesforecasting problems having said that still some of our students are confused on how to apply these techniques to real-world problems because all these concepts are spread around the course they are not all discussed at one place because it's not possible to do it to understand illicium so we have to cover all of deep learning but to do simple weighted moving averages and simple concepts basic concepts are good enough right to understand Fourier transforms a little bit of mathematical maturity is required so these concepts are spread around the whole course as and when the need arise this is slightly confusing for our students because the little cart of where a real-world problem comes up they're little caught up on what technique to apply which technique is most relevant here and things like that so what we thought we'll do is we'll take some real-world cases and in each of these cases we'll discuss which techniques to apply how to apply them gracefully and I will give pointers to the course videos where the concept has been discussed in lot more detail having said that there are some classical techniques like ARIMA that we have not discussed in the course because today state-of-the-art in forecasting is all in lists iums but what I'll try to do in this live session as time permits we'll try and cover a few classical techniques also so that your breadth of knowledge as well as your your your conceptual understanding of alternative techniques also improves right so what we'll do in this two-hour session is we'll try to take a bunch of problems we will try to solve them using techniques that we've already discussed in the course videos at the same time as and when there is something or as and when there is a context where we can introduce a new concept or a new idea or a new method or technique we will discuss that in detail again we will try to finish all of this in to us in case we run out of time we'll do one more live session on this right but I am hoping that because a lot of this content is already covered in the course videos I am hoping that we should be able to cover it in to us if there is any lack of time if there is if there is it for any reason we are not able to cover everything that we plan to cover we can do one more live session on this topic right having said this this is a this live session can be beneficial to students who are both in the early sections of the course videos as well as in the later sections of the course videos of course students who are in the earlier sections of the course videos may not be able to understand when I talk about ELISA tiems etc but they get a much more much more deeper understanding and contextual understanding on where ELISA tiems are actually useful in the bigger picture of things right but we'll be using again I'll be referring to very simple basic data analysis and basic concepts which even students were in the earlier sections of the course can benefit from certainly students who are in the later sections of the course will get a much better picture because they would have already learned about LST M's they would have done all the case studies in the course and things like that right so we will try to do an X as extensive as possible live session on time series analysis if you cannot do it in one will do it in two live sessions okay see you this Sunday hoping to have a great session with all of youhi friends the next life session that we will be conducting will be on the 5th of Jan which is the coming Sunday from 10:00 a.m. to 12:00 in the noon and this time the topic is time series analysis and forecasting this will be accessible to all of our registered students why the desktop app and as usual will you slack for all of the chat during the or any live communication during the during the live session itself having said that we have covered lots of related topics around time series analysis and forecasting the course videos itself for example we have discussed some very simple stuff in our taxi in a taxi demand prediction problem which is which is one of our case studies where we talk about simple concepts like weighted moving averages or exponentially weighted moving averages we also look at how to pause a time series forecasting problem as a typical regression problem again we also discuss about 40's Fourier transforms or Fourier series on how to detect the repeated Nisour the odd-odd or the or at what intervals is a is the time series repeating we have used some of these concepts extensively in the case study of taxi demand prediction also we also have discussed illicit iums in lots of detail which are almost the state of the art for all of timeseriesforecasting problems having said that still some of our students are confused on how to apply these techniques to real-world problems because all these concepts are spread around the course they are not all discussed at one place because it's not possible to do it to understand illicium so we have to cover all of deep learning but to do simple weighted moving averages and simple concepts basic concepts are good enough right to understand Fourier transforms a little bit of mathematical maturity is required so these concepts are spread around the whole course as and when the need arise this is slightly confusing for our students because the little cart of where a real-world problem comes up they're little caught up on what technique to apply which technique is most relevant here and things like that so what we thought we'll do is we'll take some real-world cases and in each of these cases we'll discuss which techniques to apply how to apply them gracefully and I will give pointers to the course videos where the concept has been discussed in lot more detail having said that there are some classical techniques like ARIMA that we have not discussed in the course because today state-of-the-art in forecasting is all in lists iums but what I'll try to do in this live session as time permits we'll try and cover a few classical techniques also so that your breadth of knowledge as well as your your your conceptual understanding of alternative techniques also improves right so what we'll do in this two-hour session is we'll try to take a bunch of problems we will try to solve them using techniques that we've already discussed in the course videos at the same time as and when there is something or as and when there is a context where we can introduce a new concept or a new idea or a new method or technique we will discuss that in detail again we will try to finish all of this in to us in case we run out of time we'll do one more live session on this right but I am hoping that because a lot of this content is already covered in the course videos I am hoping that we should be able to cover it in to us if there is any lack of time if there is if there is it for any reason we are not able to cover everything that we plan to cover we can do one more live session on this topic right having said this this is a this live session can be beneficial to students who are both in the early sections of the course videos as well as in the later sections of the course videos of course students who are in the earlier sections of the course videos may not be able to understand when I talk about ELISA tiems etc but they get a much more much more deeper understanding and contextual understanding on where ELISA tiems are actually useful in the bigger picture of things right but we'll be using again I'll be referring to very simple basic data analysis and basic concepts which even students were in the earlier sections of the course can benefit from certainly students who are in the later sections of the course will get a much better picture because they would have already learned about LST M's they would have done all the case studies in the course and things like that right so we will try to do an X as extensive as possible live session on time series analysis if you cannot do it in one will do it in two live sessions okay see you this Sunday hoping to have a great session with all of you\n"