Log in to your group web
You are not logged in KTH, so we cannot customize the content.
Latest posts, comments, etc in descending order
25 September at 17:38
Welcome to the following thesis defense:Student: Jakub Kowalczewski
Date and Time: 9am, Wednesday, 2 October 2019 in Ada room (Electrum building) Examiner: Šarūnas Girdzijauskas Supervisor(s): Henrik Boström Title: "Normalized conformal prediction for time series data" Opponents: José María Vera Barberán Abstract Every forecast is valid only if proper prediction intervals are stated. Currently models focus mainly on point forecast and neglect the area of prediction intervals. The estimation of the error of the model is made and is applied to every prediction in the same way, whereas we could identify that every case is different and different error measure should be applied to every instance. One of the state-of-the-art techniques which can address this behaviour is conformal prediction with its variant of normalized conformal prediction. In this thesis we will try to apply this technique into time series problems. The special focus will be put to examine the technique of estimating the difficulty of every instance using the error of neighbouring instances. This thesis will describe the entire process of adjusting time series data into normalized conformal prediction framework and the comparison with other techniques will be made.