Using d0 to Optimize Rankings
نویسندگان
چکیده
We introduce d0 as an optimization measure for ranking problems, such as information retrieval (IR). d0 is an intuitive, differentiable measure which is somewhat correlated with standard IR measures of performance. A set of experiments is described in which the weight used to combine the results of two IR systems is chosen by optimizing d0. This technique works about as well as optimizing “exact precision,” a traditional IR performance measure. These results suggest the use of d0 for training more complex, neural net models for solving ranking problems.
منابع مشابه
ILPS at TREC 2017 Common Core Track
The TREC 2017 Common Core Track aimed at gathering a diverse set of participating runs and building a new test collection using advanced pooling methods. In this paper, we describe the participation of the IlpsUvA team at the TREC 2017 Common Core Track. We submitted runs created using two methods to the track: (1) BOIR uses Bayesian optimization to automatically optimize retrieval model hyperp...
متن کاملویژگیهای نیم‌فلزی و پایداری ساختاری ترکیبات نیم‌هویسلر d0-d XYBi (X=K, Rb; Y=Sc, Ti, V, Cr)
با استفاده از محاسبات اصول اولیة ساختار الکترونی در چارچوب نظریة تابعی چگالی، ویژگیهای الکترونی، مغناطیسی و پایداری ساختاری ترکیبات نیم‌هویسلر d0-d جدید XYBi (X=K, Rb; Y=Sc, Ti, V, Cr) مورد بررسی قرار گرفت. نتایج نشان می‌دهد که ۶ مورد از این ترکیبات شاملKCrBi ،RbVBi ،KVBi ،RbTiBi ،KTiBi و RbCrBi فرومغناطیس‌های نیم-فلزی می‌باشند. گاف انرژی نیم‌فلزی این ترکیبات در محدودة ۴۶/۰ تا ۷۱/۰ الکترون‌ولت...
متن کاملClassifying Quadrics using Exact Arithmetic
2.1 The Case All di 6= 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.2 The Case d0 = 0 < d1 ≤ d2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.3 The Case d0 < 0 = d1 < d2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.4 The Case d0 ≤ d1 < d2 = 0 . . . . . . . . . . . . . . . . . . . . . . . ...
متن کاملMulti-Agent Influence Diagrams for Representing and Solving Games
The traditional representations of games using the extensive form or the strategic (normal) form obscure much of the structure that is present in real-world games. In this paper, we propose a new representation language for general multiplayer games — multi-agent influence diagrams (MAIDs). This representation extends graphical models for probability distributions to a multi-agent decision-maki...
متن کاملHeterogeneous nuclear ribonucleoprotein D0 contains transactivator and DNA-binding domains.
Heterogeneous nuclear ribonucleoprotein D0 (hnRNP D0) is an abundant, ubiquitous protein that binds RNA and DNA sequences specifically, and has been implicated in the transcriptional regulation of the human complement receptor 2 gene. We found that in vivo expression of hnRNP D0-GAL4 fusion proteins increased the transcriptional activity of a GAL4-driven reporter gene, providing direct proof th...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1998