Search (7 results, page 1 of 1)

0.031122928 = product of:
  0.062245857 = sum of:
    0.062245857 = product of:
      0.124491714 = sum of:
        0.124491714 = weight(_text_:t in 3887) [ClassicSimilarity], result of:
          0.124491714 = score(doc=3887,freq=8.0), product of:
            0.17876579 = queryWeight, product of:
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.04537884 = queryNorm
            0.69639564 = fieldWeight in 3887, product of:
              2.828427 = tf(freq=8.0), with freq of:
                8.0 = termFreq=8.0
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.0625 = fieldNorm(doc=3887)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

Although extremely useful for visualizing high-dimensional data, t-SNE plots can sometimes be mysterious or misleading. By exploring how it behaves in simple cases, we can learn to use it more effectively. We'll walk through a series of simple examples to illustrate what t-SNE diagrams can and cannot show. The t-SNE technique really is useful-but only if you know how to interpret it.

0.03044693 = product of:
  0.06089386 = sum of:
    0.06089386 = product of:
      0.12178772 = sum of:
        0.12178772 = weight(_text_:t in 3886) [ClassicSimilarity], result of:
          0.12178772 = score(doc=3886,freq=10.0), product of:
            0.17876579 = queryWeight, product of:
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.04537884 = queryNorm
            0.6812697 = fieldWeight in 3886, product of:
              3.1622777 = tf(freq=10.0), with freq of:
                10.0 = termFreq=10.0
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.0546875 = fieldNorm(doc=3886)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

The paper investigates the acceleration of t-SNE-an embedding technique that is commonly used for the visualization of high-dimensional data in scatter plots-using two tree-based algorithms. In particular, the paper develops variants of the Barnes-Hut algorithm and of the dual-tree algorithm that approximate the gradient used for learning t-SNE embeddings in O(N*logN). Our experiments show that the resulting algorithms substantially accelerate t-SNE, and that they make it possible to learn embeddings of data sets with millions of objects. Somewhat counterintuitively, the Barnes-Hut variant of t-SNE appears to outperform the dual-tree variant.

0.027232561 = product of:
  0.054465123 = sum of:
    0.054465123 = product of:
      0.108930245 = sum of:
        0.108930245 = weight(_text_:t in 3883) [ClassicSimilarity], result of:
          0.108930245 = score(doc=3883,freq=8.0), product of:
            0.17876579 = queryWeight, product of:
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.04537884 = queryNorm
            0.60934615 = fieldWeight in 3883, product of:
              2.828427 = tf(freq=8.0), with freq of:
                8.0 = termFreq=8.0
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.0546875 = fieldNorm(doc=3883)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

The paper presents a new unsupervised dimensionality reduction technique, called parametric t-SNE, that learns a parametric mapping between the high-dimensional data space and the low-dimensional latent space. Parametric t-SNE learns the parametric mapping in such a way that the local structure of the data is preserved as well as possible in the latent space. We evaluate the performance of parametric t-SNE in experiments on three datasets, in which we compare it to the performance of two other unsupervised parametric dimensionality reduction techniques. The results of experiments illustrate the strong performance of parametric t-SNE, in particular, in learning settings in which the dimensionality of the latent space is relatively low.

0.02382353 = product of:
  0.04764706 = sum of:
    0.04764706 = product of:
      0.09529412 = sum of:
        0.09529412 = weight(_text_:t in 3888) [ClassicSimilarity], result of:
          0.09529412 = score(doc=3888,freq=12.0), product of:
            0.17876579 = queryWeight, product of:
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.04537884 = queryNorm
            0.53306687 = fieldWeight in 3888, product of:
              3.4641016 = tf(freq=12.0), with freq of:
                12.0 = termFreq=12.0
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.0390625 = fieldNorm(doc=3888)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

We present a new technique called "t-SNE" that visualizes high-dimensional data by giving each datapoint a location in a two or three-dimensional map. The technique is a variation of Stochastic Neighbor Embedding (Hinton and Roweis, 2002) that is much easier to optimize, and produces significantly better visualizations by reducing the tendency to crowd points together in the center of the map. t-SNE is better than existing techniques at creating a single map that reveals structure at many different scales. This is particularly important for high-dimensional data that lie on several different, but related, low-dimensional manifolds, such as images of objects from multiple classes seen from multiple viewpoints. For visualizing the structure of very large data sets, we show how t-SNE can use random walks on neighborhood graphs to allow the implicit structure of all of the data to influence the way in which a subset of the data is displayed. We illustrate the performance of t-SNE on a wide variety of data sets and compare it with many other non-parametric visualization techniques, including Sammon mapping, Isomap, and Locally Linear Embedding. The visualizations produced by t-SNE are significantly better than those produced by the other techniques on almost all of the data sets.

0.020214936 = product of:
  0.04042987 = sum of:
    0.04042987 = product of:
      0.08085974 = sum of:
        0.08085974 = weight(_text_:t in 3884) [ClassicSimilarity], result of:
          0.08085974 = score(doc=3884,freq=6.0), product of:
            0.17876579 = queryWeight, product of:
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.04537884 = queryNorm
            0.45232224 = fieldWeight in 3884, product of:
              2.4494898 = tf(freq=6.0), with freq of:
                6.0 = termFreq=6.0
              3.9394085 = idf(docFreq=2338, maxDocs=44218)
              0.046875 = fieldNorm(doc=3884)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

Techniques for multidimensional scaling visualize objects as points in a low-dimensional metric map. As a result, the visualizations are subject to the fundamental limitations of metric spaces. These limitations prevent multidimensional scaling from faithfully representing non-metric similarity data such as word associations or event co-occurrences. In particular, multidimensional scaling cannot faithfully represent intransitive pairwise similarities in a visualization, and it cannot faithfully visualize "central" objects. In this paper, we present an extension of a recently proposed multidimensional scaling technique called t-SNE. The extension aims to address the problems of traditional multidimensional scaling techniques when these techniques are used to visualize non-metric similarities. The new technique, called multiple maps t-SNE, alleviates these problems by constructing a collection of maps that reveal complementary structure in the similarity data. We apply multiple maps t-SNE to a large data set of word association data and to a data set of NIPS co-authorships, demonstrating its ability to successfully visualize non-metric similarities.

0.012296414 = product of:
  0.024592828 = sum of:
    0.024592828 = product of:
      0.049185656 = sum of:
        0.049185656 = weight(_text_:22 in 4261) [ClassicSimilarity], result of:
          0.049185656 = score(doc=4261,freq=2.0), product of:
            0.15890898 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.04537884 = queryNorm
            0.30952093 = fieldWeight in 4261, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.0625 = fieldNorm(doc=4261)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Date

17. 7.2002 19:22:06

0.006148207 = product of:
  0.012296414 = sum of:
    0.012296414 = product of:
      0.024592828 = sum of:
        0.024592828 = weight(_text_:22 in 5234) [ClassicSimilarity], result of:
          0.024592828 = score(doc=5234,freq=2.0), product of:
            0.15890898 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.04537884 = queryNorm
            0.15476047 = fieldWeight in 5234, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03125 = fieldNorm(doc=5234)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Date

22. 1.2018 11:33:49