Conference Proceeding
Refine
Year of publication
Institute
- Fachbereich Wirtschaftswissenschaften (73) (remove)
Document Type
- Conference Proceeding (73) (remove)
Keywords
- Telekommunikationsmarkt (6)
- Führung (4)
- Leadership (4)
- Motivation (2)
- Self-Leadership (2)
- regulation (2)
- Active learning (1)
- Breitband Markt (1)
- Bundesnetzagentur (1)
- Business Process Intelligence (1)
Supervised machine learning and deep learning require a large amount of labeled data, which data scientists obtain in a manual, and time-consuming annotation process. To mitigate this challenge, Active Learning (AL) proposes promising data points to annotators they annotate next instead of a subsequent or random sample. This method is supposed to save annotation effort while maintaining model performance.
However, practitioners face many AL strategies for different tasks and need an empirical basis to choose between them. Surveys categorize AL strategies into taxonomies without performance indications. Presentations of novel AL strategies compare the performance to a small subset of strategies. Our contribution addresses the empirical basis by introducing a reproducible active learning evaluation (ALE) framework for the comparative evaluation of AL strategies in NLP.
The framework allows the implementation of AL strategies with low effort and a fair data-driven comparison through defining and tracking experiment parameters (e.g., initial dataset size, number of data points per query step, and the budget). ALE helps practitioners to make more informed decisions, and researchers can focus on developing new, effective AL strategies and deriving best practices for specific use cases. With best practices, practitioners can lower their annotation costs. We present a case study to illustrate how to use the framework.
Leveraging Social Network Data for Analytical CRM Strategies - The Introduction of Social BI.
(2012)
Adaptive logistics : information management for planning and control of small series assembly
(2007)
Bitcoin is a cryptocurrency and is considered a high-risk asset class whose price changes are difficult to predict. Current research focusses on daily price movements with a limited number of predictors. The paper at hand aims at identifying measurable indicators for Bitcoin price movements and the development of a suitable forecasting model for hourly changes. The paper provides three research contributions. First, a set of significant indicators for predicting the Bitcoin price is identified. Second, the results of a trained Long Short-term Memory (LSTM) neural network that predicts price changes on an hourly basis is presented and compared with other algorithms. Third, the results foster discussions of the applicability of neural nets for stock price predictions. In total, 47 input features for a period of over 10 months could be retrieved to train a neural net that predicts the Bitcoin price movements with an error rate of 3.52 %.
Concept, scientific research and managerial applications of Provocative Coaching, according to the „Provocative Therapy“ of Prof. Dr. Frank Farrelly (University of Wisconsin, U.S.A) in terms of an application of the Provocative Communication Style in specific situations of practical leadership, especially in the role of a coach for their subordinates.