Document (#40888)

Author

Wattenberg, M.
Viégas, F.
Johnson, I.

Title

How to use t-SNE effectively

Source

Distill, [http://doi.org/10.23915/distill.00002]

Year

2016

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.

Content

Vgl.: https://distill.pub/2016/misread-tsne/.

Theme

Data Mining
Visualisierung

Object

tSNE

Similar documents (author)

1. Johnson, S.W.: Do-it-yourself CD-ROMs (1992) 4.57

   4.566886 = sum of:
     4.566886 = weight(author_txt:johnson in 4284) [ClassicSimilarity], result of:
       4.566886 = score(doc=4284,freq=1.0), product of:
         0.99999994 = queryWeight, product of:
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.13685472 = queryNorm
         4.5668864 = fieldWeight in 4284, product of:
           1.0 = tf(freq=1.0), with freq of:
             1.0 = termFreq=1.0
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.625 = fieldNorm(doc=4284)
   

2. Johnson, S.: Virtual documents : the past, the present and some standards for the future (1993) 4.57

   4.566886 = sum of:
     4.566886 = weight(author_txt:johnson in 4420) [ClassicSimilarity], result of:
       4.566886 = score(doc=4420,freq=1.0), product of:
         0.99999994 = queryWeight, product of:
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.13685472 = queryNorm
         4.5668864 = fieldWeight in 4420, product of:
           1.0 = tf(freq=1.0), with freq of:
             1.0 = termFreq=1.0
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.625 = fieldNorm(doc=4420)
   

3. Johnson, R.D.: Public libraries and the Internet / NREN : new challenges, new opportunities (1992) 4.57

   4.566886 = sum of:
     4.566886 = weight(author_txt:johnson in 6247) [ClassicSimilarity], result of:
       4.566886 = score(doc=6247,freq=1.0), product of:
         0.99999994 = queryWeight, product of:
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.13685472 = queryNorm
         4.5668864 = fieldWeight in 6247, product of:
           1.0 = tf(freq=1.0), with freq of:
             1.0 = termFreq=1.0
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.625 = fieldNorm(doc=6247)
   

4. Johnson, F.C.: ¬A classification of ellipsis based on a corpus of information seeking dialogues (1994) 4.57

   4.566886 = sum of:
     4.566886 = weight(author_txt:johnson in 7802) [ClassicSimilarity], result of:
       4.566886 = score(doc=7802,freq=1.0), product of:
         0.99999994 = queryWeight, product of:
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.13685472 = queryNorm
         4.5668864 = fieldWeight in 7802, product of:
           1.0 = tf(freq=1.0), with freq of:
             1.0 = termFreq=1.0
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.625 = fieldNorm(doc=7802)
   

5. Johnson, A.: Information brokers (1991) 4.57

   4.566886 = sum of:
     4.566886 = weight(author_txt:johnson in 1362) [ClassicSimilarity], result of:
       4.566886 = score(doc=1362,freq=1.0), product of:
         0.99999994 = queryWeight, product of:
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.13685472 = queryNorm
         4.5668864 = fieldWeight in 1362, product of:
           1.0 = tf(freq=1.0), with freq of:
             1.0 = termFreq=1.0
           7.3070183 = idf(docFreq=80, maxDocs=44421)
           0.625 = fieldNorm(doc=1362)
   

Similar documents (content)

1. Maaten, L. van den: Accelerating t-SNE using Tree-Based Algorithms (2014) 0.12

   0.116099715 = sum of:
     0.116099715 = product of:
       0.5804986 = sum of:
         0.041470096 = weight(abstract_txt:high in 4886) [ClassicSimilarity], result of:
           0.041470096 = score(doc=4886,freq=1.0), product of:
             0.09120039 = queryWeight, product of:
               1.0145345 = boost
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.01853373 = queryNorm
             0.454714 = fieldWeight in 4886, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.09375 = fieldNorm(doc=4886)
         0.063636295 = weight(abstract_txt:technique in 4886) [ClassicSimilarity], result of:
           0.063636295 = score(doc=4886,freq=1.0), product of:
             0.121332355 = queryWeight, product of:
               1.170191 = boost
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.01853373 = queryNorm
             0.5244792 = fieldWeight in 4886, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.09375 = fieldNorm(doc=4886)
         0.09005467 = weight(abstract_txt:learn in 4886) [ClassicSimilarity], result of:
           0.09005467 = score(doc=4886,freq=1.0), product of:
             0.1529364 = queryWeight, product of:
               1.3137838 = boost
               6.2809324 = idf(docFreq=225, maxDocs=44421)
               0.01853373 = queryNorm
             0.5888374 = fieldWeight in 4886, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.2809324 = idf(docFreq=225, maxDocs=44421)
               0.09375 = fieldNorm(doc=4886)
         0.11483896 = weight(abstract_txt:dimensional in 4886) [ClassicSimilarity], result of:
           0.11483896 = score(doc=4886,freq=1.0), product of:
             0.17984548 = queryWeight, product of:
               1.424683 = boost
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.01853373 = queryNorm
             0.63854235 = fieldWeight in 4886, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.09375 = fieldNorm(doc=4886)
         0.2704986 = weight(abstract_txt:plots in 4886) [ClassicSimilarity], result of:
           0.2704986 = score(doc=4886,freq=1.0), product of:
             0.31838313 = queryWeight, product of:
               1.895586 = boost
               9.06241 = idf(docFreq=13, maxDocs=44421)
               0.01853373 = queryNorm
             0.849601 = fieldWeight in 4886, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               9.06241 = idf(docFreq=13, maxDocs=44421)
               0.09375 = fieldNorm(doc=4886)
       0.2 = coord(5/25)
   

2. Maaten, L. van den; Hinton, G.: Visualizing data using t-SNE (2008) 0.08

   0.08238038 = sum of:
     0.08238038 = product of:
       0.4119019 = sum of:
         0.03909838 = weight(abstract_txt:high in 4888) [ClassicSimilarity], result of:
           0.03909838 = score(doc=4888,freq=2.0), product of:
             0.09120039 = queryWeight, product of:
               1.0145345 = boost
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.01853373 = queryNorm
             0.42870846 = fieldWeight in 4888, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.0625 = fieldNorm(doc=4888)
         0.059996873 = weight(abstract_txt:technique in 4888) [ClassicSimilarity], result of:
           0.059996873 = score(doc=4888,freq=2.0), product of:
             0.121332355 = queryWeight, product of:
               1.170191 = boost
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.01853373 = queryNorm
             0.4944837 = fieldWeight in 4888, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.0625 = fieldNorm(doc=4888)
         0.053721193 = weight(abstract_txt:illustrate in 4888) [ClassicSimilarity], result of:
           0.053721193 = score(doc=4888,freq=1.0), product of:
             0.14201404 = queryWeight, product of:
               1.2660012 = boost
               6.0524936 = idf(docFreq=283, maxDocs=44421)
               0.01853373 = queryNorm
             0.37828085 = fieldWeight in 4888, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.0524936 = idf(docFreq=283, maxDocs=44421)
               0.0625 = fieldNorm(doc=4888)
         0.15311861 = weight(abstract_txt:dimensional in 4888) [ClassicSimilarity], result of:
           0.15311861 = score(doc=4888,freq=4.0), product of:
             0.17984548 = queryWeight, product of:
               1.424683 = boost
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.01853373 = queryNorm
             0.8513898 = fieldWeight in 4888, product of:
               2.0 = tf(freq=4.0), with freq of:
                 4.0 = termFreq=4.0
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.0625 = fieldNorm(doc=4888)
         0.10596683 = weight(abstract_txt:visualizing in 4888) [ClassicSimilarity], result of:
           0.10596683 = score(doc=4888,freq=1.0), product of:
             0.2233645 = queryWeight, product of:
               1.5877259 = boost
               7.590594 = idf(docFreq=60, maxDocs=44421)
               0.01853373 = queryNorm
             0.4744121 = fieldWeight in 4888, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               7.590594 = idf(docFreq=60, maxDocs=44421)
               0.0625 = fieldNorm(doc=4888)
       0.2 = coord(5/25)
   

3. Hochheiser, H.; Shneiderman, B.: Using interactive visualizations of WWW log data to characterize access patterns and inform site design (2001) 0.08

   0.07782864 = sum of:
     0.07782864 = product of:
       0.38914317 = sum of:
         0.03309431 = weight(abstract_txt:although in 6765) [ClassicSimilarity], result of:
           0.03309431 = score(doc=6765,freq=1.0), product of:
             0.088605985 = queryWeight, product of:
               4.780796 = idf(docFreq=1012, maxDocs=44421)
               0.01853373 = queryNorm
             0.3734997 = fieldWeight in 6765, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.780796 = idf(docFreq=1012, maxDocs=44421)
               0.078125 = fieldNorm(doc=6765)
         0.050304685 = weight(abstract_txt:series in 6765) [ClassicSimilarity], result of:
           0.050304685 = score(doc=6765,freq=1.0), product of:
             0.117138535 = queryWeight, product of:
               1.1497896 = boost
               5.4969096 = idf(docFreq=494, maxDocs=44421)
               0.01853373 = queryNorm
             0.42944607 = fieldWeight in 6765, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.4969096 = idf(docFreq=494, maxDocs=44421)
               0.078125 = fieldNorm(doc=6765)
         0.09569913 = weight(abstract_txt:dimensional in 6765) [ClassicSimilarity], result of:
           0.09569913 = score(doc=6765,freq=1.0), product of:
             0.17984548 = queryWeight, product of:
               1.424683 = boost
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.01853373 = queryNorm
             0.5321186 = fieldWeight in 6765, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.8111186 = idf(docFreq=132, maxDocs=44421)
               0.078125 = fieldNorm(doc=6765)
         0.11312279 = weight(abstract_txt:interpret in 6765) [ClassicSimilarity], result of:
           0.11312279 = score(doc=6765,freq=1.0), product of:
             0.20106089 = queryWeight, product of:
               1.5063721 = boost
               7.201658 = idf(docFreq=89, maxDocs=44421)
               0.01853373 = queryNorm
             0.5626295 = fieldWeight in 6765, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               7.201658 = idf(docFreq=89, maxDocs=44421)
               0.078125 = fieldNorm(doc=6765)
         0.09692225 = weight(abstract_txt:useful in 6765) [ClassicSimilarity], result of:
           0.09692225 = score(doc=6765,freq=2.0), product of:
             0.18137462 = queryWeight, product of:
               2.0233533 = boost
               4.83662 = idf(docFreq=957, maxDocs=44421)
               0.01853373 = queryNorm
             0.534376 = fieldWeight in 6765, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               4.83662 = idf(docFreq=957, maxDocs=44421)
               0.078125 = fieldNorm(doc=6765)
       0.2 = coord(5/25)
   

4. Schamber, L.: Time-line interviews and inductive content analysis : their effectiveness for exploring cognitive behaviors (2000) 0.07

   0.06523129 = sum of:
     0.06523129 = product of:
       0.4076956 = sum of:
         0.052637305 = weight(abstract_txt:examples in 5808) [ClassicSimilarity], result of:
           0.052637305 = score(doc=5808,freq=1.0), product of:
             0.09647273 = queryWeight, product of:
               1.0434479 = boost
               4.9885116 = idf(docFreq=822, maxDocs=44421)
               0.01853373 = queryNorm
             0.5456185 = fieldWeight in 5808, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.9885116 = idf(docFreq=822, maxDocs=44421)
               0.109375 = fieldNorm(doc=5808)
         0.10146612 = weight(abstract_txt:exploring in 5808) [ClassicSimilarity], result of:
           0.10146612 = score(doc=5808,freq=1.0), product of:
             0.14942487 = queryWeight, product of:
               1.2986134 = boost
               6.208406 = idf(docFreq=242, maxDocs=44421)
               0.01853373 = queryNorm
             0.6790444 = fieldWeight in 5808, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.208406 = idf(docFreq=242, maxDocs=44421)
               0.109375 = fieldNorm(doc=5808)
         0.15764403 = weight(abstract_txt:extremely in 5808) [ClassicSimilarity], result of:
           0.15764403 = score(doc=5808,freq=1.0), product of:
             0.20044437 = queryWeight, product of:
               1.5040609 = boost
               7.190608 = idf(docFreq=90, maxDocs=44421)
               0.01853373 = queryNorm
             0.78647274 = fieldWeight in 5808, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               7.190608 = idf(docFreq=90, maxDocs=44421)
               0.109375 = fieldNorm(doc=5808)
         0.09594814 = weight(abstract_txt:useful in 5808) [ClassicSimilarity], result of:
           0.09594814 = score(doc=5808,freq=1.0), product of:
             0.18137462 = queryWeight, product of:
               2.0233533 = boost
               4.83662 = idf(docFreq=957, maxDocs=44421)
               0.01853373 = queryNorm
             0.5290053 = fieldWeight in 5808, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               4.83662 = idf(docFreq=957, maxDocs=44421)
               0.109375 = fieldNorm(doc=5808)
       0.16 = coord(4/25)
   

5. Costa Carvalho, A. da; Rossi, C.; Moura, E.S. de; Silva, A.S. da; Fernandes, D.: LePrEF: Learn to precompute evidence fusion for efficient query evaluation (2012) 0.06

   0.062912755 = sum of:
     0.062912755 = product of:
       0.31456375 = sum of:
         0.03909838 = weight(abstract_txt:high in 1278) [ClassicSimilarity], result of:
           0.03909838 = score(doc=1278,freq=2.0), product of:
             0.09120039 = queryWeight, product of:
               1.0145345 = boost
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.01853373 = queryNorm
             0.42870846 = fieldWeight in 1278, product of:
               1.4142135 = tf(freq=2.0), with freq of:
                 2.0 = termFreq=2.0
               4.8502827 = idf(docFreq=944, maxDocs=44421)
               0.0625 = fieldNorm(doc=1278)
         0.0424242 = weight(abstract_txt:technique in 1278) [ClassicSimilarity], result of:
           0.0424242 = score(doc=1278,freq=1.0), product of:
             0.121332355 = queryWeight, product of:
               1.170191 = boost
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.01853373 = queryNorm
             0.3496528 = fieldWeight in 1278, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.5944448 = idf(docFreq=448, maxDocs=44421)
               0.0625 = fieldNorm(doc=1278)
         0.06003645 = weight(abstract_txt:learn in 1278) [ClassicSimilarity], result of:
           0.06003645 = score(doc=1278,freq=1.0), product of:
             0.1529364 = queryWeight, product of:
               1.3137838 = boost
               6.2809324 = idf(docFreq=225, maxDocs=44421)
               0.01853373 = queryNorm
             0.39255828 = fieldWeight in 1278, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               6.2809324 = idf(docFreq=225, maxDocs=44421)
               0.0625 = fieldNorm(doc=1278)
         0.07274137 = weight(abstract_txt:simple in 1278) [ClassicSimilarity], result of:
           0.07274137 = score(doc=1278,freq=1.0), product of:
             0.2189938 = queryWeight, product of:
               2.2233067 = boost
               5.314588 = idf(docFreq=593, maxDocs=44421)
               0.01853373 = queryNorm
             0.33216175 = fieldWeight in 1278, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.314588 = idf(docFreq=593, maxDocs=44421)
               0.0625 = fieldNorm(doc=1278)
         0.10026336 = weight(abstract_txt:effectively in 1278) [ClassicSimilarity], result of:
           0.10026336 = score(doc=1278,freq=1.0), product of:
             0.27123082 = queryWeight, product of:
               2.4743035 = boost
               5.9145703 = idf(docFreq=325, maxDocs=44421)
               0.01853373 = queryNorm
             0.36966065 = fieldWeight in 1278, product of:
               1.0 = tf(freq=1.0), with freq of:
                 1.0 = termFreq=1.0
               5.9145703 = idf(docFreq=325, maxDocs=44421)
               0.0625 = fieldNorm(doc=1278)
       0.2 = coord(5/25)