To forecast the movement directions of stocks, exchange rates, and stock markets are significant and an active research area for investors, analysts, and researchers. In this paper, word embedding and deep learning-based direction prediction of Istanbul Stock Exchange (BIST 100) is proposed by analyzing nine banking stocks with high volume in BIST 100. Though English news articles have been employed for forecasting of market direction previously, to the best of our knowledge, Turkish news articles and user comments from social media and different platforms have not been utilized with the combination of deep learning techniques and word embedding methods to predict the direction of Turkish stocks and market. For this objective, long short-term memory networks, recurrent neural networks, convolutional neural networks as deep learning algorithms and Word2Vec, GloVe, and FastText as word embedding models are evaluated. To demonstrate the effectiveness of proposed model, four different sources of Turkish news are collected. The news articles about stocks from Public Disclosure Platform (KAP), text-based technical analysis of each stock from Bigpara, user comments from both Twitter and Mynet Finans platforms are gathered. Experiment results demonstrate that the combination of deep learning techniques and word embedding methods have a great potential to predict the direction of BIST 100.
Robots and artificial intelligence technologies have become very important in the health applications as in many other fields. The proposed system in this work aims to provide detailed analysis of pre-op and post-op stage of FUE hair transplant procedures to enable surgeon to plan and assess success of the operations. In order to achieve this target, a robotic and vision-based system imaging and AI based analysis approach is developed. The proposed system performs analyses in three main stages: initialization, scanning, and analysis. At the initialization stage, 3D model of the patient’s head generated at first by locating a depth camera in various positions around the patient by the help of a collaborative robot. At the second stage, where high resolution image capturing is performed in a loop with the usage of the 3D model, raw images are processed by a deep learning based object detection algorithm where follicles in pre-op and extracted follicle positions (i.e. holes) and placed grafts in post-op is detected. At the last stage, thickness of each hair is computed at the detected hair follicle positions using another deep learning-based segmentation approach. These data are combined to obtain objective evaluation criteria to generate patient report. Experimental results show that the developed system can be used successfully in hair transplantation operations.
21 August 2019 in Signal Processing: Image Communication
Bu çalışmada düşük bit derinliği gösterimi temeli hareket kestirimi yaklaşımlarının en güncel video kodlama standardı olan HEVC üzerinde detaylı bir karşılaştırmalı analiz gerçekleştirilmiştir. Böylelikle geçmişte çoğunlukla açık çevrim performans değerlendirmesine tabii tutulan bu yaklaşımların bir kodlayıcı içerisindeki gerçek performansı ortaya çıkarılmıştır. Bunun yanı sıra, düşük bit‑derinliği gösterimi temelli hareket kestirim yaklaşımlarının daha da hızlandırılması için seyrek arama yaklaşımları ile tümleştirerek HEVC üzerine uygulanması literatürde ilk kez incelenmiştir. Bu inceleme sonucunda HEVC referans yazılımında kullanılan TZS isimli algoritmanın düşük-bit derinliğindeki görüntüler üzerinde kullanımı durumunda performans kaybı oluşacağı çıkarılmıştır.
High classification scores are meaningful only if the prediction target is physically consistent. We audited image-only, tabular-only and fusion experiments on 9000 ZooScan objects drawn from the 1,153,507-object PELGAS collection. The archived three-class labels were tertiles of the stored equivalent spherical diameter (ESD), not biomass classes. All 893,542 records exported with version v7.26 followed the documented ESD formula, whereas all 259,965 v7.24 records equalled twice the integer part of projected area divided by . All 2017 v7.24 objects in the subset fell in the highest archived class. Recomputing ESD consistently from area and pixel calibration changed 3430 of the 9000 labels. The subset spans 13 cruises and 697 cruise–station groups and includes 2639 objects annotated as non-living. On the corrected target, tree-based controls using all descriptors reached a macro-F1 of about 0.998 under both object-stratified and station-grouped evaluation. In preliminary single-seed neural runs without area- or length-based descriptors, tabular networks reached a macro-F1 of 0.89, a Swin-T image model 0.946 and gated fusion 0.944; the image-based runs covered three folds and retained scale cues, so they do not isolate the contribution of each modality. Target provenance should be verified before multimodal performance rankings are used in ecological monitoring.