Time Series - 8 Non-linear Trend - Exponential Curve Fitting

[@teresiahnganga948]

This has been very helpful ☺️

[@PUAARAcademy]

Glad you think so! Keep watching, keep learning... Please don't forget to like and share... And also visit, subscribe and recommend my channel kzbin.info/door/qedkv6AYxh3hk99EPtGlmQ

[@chengningloong7691]

Sir, I have a problem: Let's say we have a time-series dataset D I separate D into training set Tr and testing set Ts (i.e. D={Tr, Ts}) The Tr consists of the first 100 timestep data obtained from experiment I use this Tr to train a model, say exponential model yt_hat=A^t+B (I obtained the optimal A and B from the training set) yt_hat is the estimated value at time t In order to test the performance of the model, I predict the yt_hat predicted from the hypothetical model and the yt from the testing set. What kind of statistical testing should I use to show that my model is accurate enough?

[@shayanlusiff]

How 5 came in Y2 interval

[@verloren2677]

Sir can you do this sum using direct method

[@PUAARAcademy]

Yes, I can do it personally for me, but I am not going to upload any lecture on it because it is useful to the few students going for specialization. I would rather prefer to upload here the lectures on the topics which are common to all the students studying various subjects in different courses at different places... :)

[@verloren2677]

Prashant Puaar It's Okay Sir and your videos and very helpful 👍👍👍👍👍👍👍👍👍👍👍

[@cdavid943]

sir for logY 12.9 it's 1.1186 not 1106

[@PUAARAcademy]

No, 1.1106 is correct. It SHOULD NOT BE 1.1186. You need to check the value in the log-table carefully (Row of 12 and column of 9). 1186 is possible only if the Y-value is 13.14 instead of 12.9. but the VALUE IS 12.9 which we can see in the 'Y' column.