Adapting Codes and Embeddings for Polychotomies

نویسندگان

  • Gunnar Rätsch
  • Alexander J. Smola
  • Sebastian Mika
چکیده

In this paper we consider formulations of multi-class problems based on a generalized notion of a margin and using output coding. This includes, but is not restricted to, standard multi-class SVM formulations. Differently from many previous approaches we learn the code as well as the embedding function. We illustrate how this can lead to a formulation that allows for solving a wider range of problems with e.g. many classes or even “missing classes”. To keep our optimization problems tractable we propose an algorithm capable of solving them using two-class classifiers, similar in spirit to Boosting.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

IDIAP Research Report 96 - 08 On the Decomposition of Polychotomies intoDichotomiesEddy

Many important classiication problems are polychotomies, i.e. the data are organized into K classes with K > 2. Given an unknown function F : ! f1; : : : ; Kg representing a polychotomy, an algorithm aimed at \learning" this polychotomy will produce an approximation of F , based on the knowledge of a set of pairs f(x p ; F (x p))g P p=1. Although in the wide variety of learning tools there exis...

متن کامل

On the Decomposition of Polychotomies into Dichotomies on the Decomposition of Polychotomies into Dichotomies

Many important classi cation problems are polychotomies i e the data are organized into K classes with K Given an unknown function F f Kg representing a polychotomy an algorithm aimed at learning this polychotomy will produce an approximation of F based on the knowledge of a set of pairs f x F x gp Although in the wide variety of learning tools there exist some learning algorithms capable of ha...

متن کامل

The pseudo-hyperplanes and homogeneous pseudo-embeddings of AG(n, 4) and PG(n, 4)

We determine all homogeneous pseudo-embeddings of the affine space AG(n, 4) and the projective space PG(n, 4). We give a classification of all pseudo-hyperplanes of AG(n, 4). We also prove that the two homogeneous pseudo-embeddings of the generalized quadrangle Q(4, 3) are induced by the two homogeneous pseudo-embeddings of AG(4, 4) into which Q(4, 3) is fully embeddable.

متن کامل

Labeling Subgraph Embeddings and Cordiality of Graphs

Let $G$ be a graph with vertex set $V(G)$ and edge set $E(G)$, a vertex labeling $f : V(G)rightarrow mathbb{Z}_2$ induces an edge labeling $ f^{+} : E(G)rightarrow mathbb{Z}_2$ defined by $f^{+}(xy) = f(x) + f(y)$, for each edge $ xyin E(G)$.  For each $i in mathbb{Z}_2$, let $ v_{f}(i)=|{u in V(G) : f(u) = i}|$ and $e_{f^+}(i)=|{xyin E(G) : f^{+}(xy) = i}|$. A vertex labeling $f$ of a graph $G...

متن کامل

Canonical Correlation Analysis for Analyzing Sequences of Medical Billing Codes

We propose using canonical correlation analysis (CCA) to generate features from sequences of medical billing codes. Applying this novel use of CCA to a database of medical billing codes for patients with diverticulitis, we first demonstrate that the CCA embeddings capture meaningful relationships among the codes. We then generate features from these embeddings and establish their usefulness in ...

متن کامل

ذخیره 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 2002