Browsing by Subject "Stacking"

Now showing 1 - 4 of 4

Open Access
GOOWE-ML: a novel online stacked ensemble for multi-label classification in data streams
(2019-07) Büyükçakır, Alican
As data streams become more prevalent, the necessity for online algorithms that mine this transient and dynamic data becomes clearer. Multi-label data stream classification is a supervised learning problem where each instance in the data stream is classified into one or more pre-defined sets of labels. Many methods have been proposed to tackle this problem, including but not limited to ensemblebased methods. Some of these ensemble-based methods are specifically designed to work with certain multi-label base classifiers; some others employ online bagging schemes to build their ensembles. In this study, we introduce a novel online and dynamically-weighted stacked ensemble for multi-label classification, called GOOWE-ML, that utilizes spatial modeling to assign optimal weights to its component classifiers. Our model can be used with any existing incremental multilabel classification algorithm as its base classifier. We conduct experiments with 4 GOOWE-ML-based multi-label ensembles and 7 baseline models on 7 real-world datasets from diverse areas of interest. Our experiments show that GOOWE-ML ensembles yield consistently better results in terms of predictive performance in almost all of the datasets, with respect to the other prominent ensemble models.
Open Access
Liquid interface self-assembly with colloidal quantum wells
(Springer Singapore, 2022-10-28) Erdem, Onur; Demir, Hilmi Volkan
In this chapter, we discuss our methodologies for the self-assembly of colloidal nanoplatelets (NPLs) at the liquid interface. We also review other recent studies on orientation-controlled platelet assembly on liquid interfaces. We compare the results of the reported studies and discuss the parameters that affect the NPL orientation at the liquid interface.
Open Access
Orientation-controlled nonradiative energy transfer to colloidal nanoplatelets: engineering dipole orientation factor
(American Chemical Society, 2019) Erdem, Onur; Güngör, Kıvanç; Güzeltürk, Burak; Tanrıöver, İbrahim; Sak, Mustafa; Olutaş, Murat; Dede, Didem; Kelestemur, Yusuf; Demir, Hilmi Volkan
We proposed and showed strongly orientation-controlled Förster resonance energy transfer (FRET) to highly anisotropic CdSe nanoplatelets (NPLs). For this purpose, we developed a liquid–air interface self-assembly technique specific to depositing a complete monolayer of NPLs only in a single desired orientation, either fully stacked (edge-up) or fully nonstacked (face-down), with near-unity surface coverage and across large areas over 20 cm2. These NPL monolayers were employed as acceptors in an energy transfer working model system to pair with CdZnS/ZnS core/shell quantum dots (QDs) as donors. We found the resulting energy transfer from the QDs to be significantly accelerated (by up to 50%) to the edge-up NPL monolayer compared to the face-down one. We revealed that this acceleration of FRET is accounted for by the enhancement of the dipole–dipole interaction factor between a QD-NPL pair (increased from 1/3 to 5/6) as well as the closer packing of NPLs with stacking. Also systematically studying the distance-dependence of FRET between QDs and NPL monolayers via varying their separation (d) with a dielectric spacer, we found out that the FRET rate scales with d–4 regardless of the specific NPL orientation. Our FRET model, which is based on the original Förster theory, computes the FRET efficiencies in excellent agreement with our experimental results and explains well the enhancement of FRET to NPLs with stacking. These findings indicate that the geometrical orientation of NPLs and thereby their dipole interaction strength can be exploited as an additional degree of freedom to control and tune the energy transfer rate.
Open Access
Ray representation for k-trees
(Elsevier, 1989) Akman, V.; Franklin, Wm. R.
k-trees have established themselves as useful data structures in pattern recognition. A fundamental operation regarding k-trees is the construction of a k-tree. We present a method to store an object as a set of rays and an algorithm to convert such a set into a k-tree. The algorithm is conceptually simple, works for any k, and builds a k-tree from the rays very fast. It produces a minimal k-tree and does not lead to intermediate storage swell. © 1989.