1mlpack_fastmks(1) User Commands mlpack_fastmks(1)
2
6 mlpack_fastmks - fastmks (fast max-kernel search)
7
9 mlpack_fastmks [-w double] [-b double] [-d double] [-m unknown] [-k int] [-K string] [-N bool] [-o double] [-q string] [-r string] [-s double] [-S bool] [-V bool] [-i string] [-p string] [-M unknown] [-h -v]
10
12 This program will find the k maximum kernels of a set of points, using
13 a query set and a reference set (which can optionally be the same set).
14 More specifically, for each point in the query set, the k points in the
15 reference set with maximum kernel evaluations are found. The kernel
16 function used is specified with the '--kernel (-K)' parameter.
17 For example, the following command will calculate, for each point in
19 the query set 'query.csv', the five points in the reference set 'refer‐
20 ence.csv' with maximum kernel evaluation using the linear kernel. The
21 kernel evaluations may be saved with the 'kernels.csv' output parameter
22 and the indices may be saved with the 'indices.csv' output parameter.
23 $ fastmks --k 5 --reference_file reference.csv --query_file query.csv
25 --indices_file indices.csv --kernels_file kernels.csv --kernel linear
26 The output matrices are organized such that row i and column j in the
28 indices matrix corresponds to the index of the point in the reference
29 set that has j'th largest kernel evaluation with the point in the query
30 set with index i. Row i and column j in the kernels matrix corresponds
31 to the kernel evaluation between those two points.
32 This program performs FastMKS using a cover tree. The base used to
34 build the cover tree can be specified with the '--base (-b)' parameter.
35
37 --bandwidth (-w) [double]
38 Bandwidth (for Gaussian, Epanechnikov, and triangular kernels).
39 Default value 1.
40 --base (-b) [double]
42 Base to use during cover tree construction. Default value 2.
43 --degree (-d) [double]
45 Degree of polynomial kernel. Default value 2.
46 --help (-h) [bool]
48 Default help info.
49 --info [string]
51 Get help on a specific module or option. Default value ''.
52 --input_model_file (-m) [unknown]
54 Input FastMKS model to use. Default value ''.
55 --k (-k) [int]
57 Number of maximum kernels to find. Default value 0.
58 --kernel (-K) [string]
60 Kernel type to use: 'linear', 'polynomial', 'cosine', 'gauss‐
61 ian', 'epanechnikov', 'triangular', 'hyptan'. Default value
62 'linear'.
63 --naive (-N) [bool]
65 If true, O(n^2) naive mode is used for computation.
66 --offset (-o) [double]
68 Offset of kernel (for polynomial and hyptan kernels). Default
69 value 0.
70 --query_file (-q) [string]
72 The query dataset. Default value ''.
73 --reference_file (-r) [string]
75 The reference dataset. Default value ''.
76 --scale (-s) [double]
78 Scale of kernel (for hyptan kernel). Default value 1.
79 --single (-S) [bool]
81 If true, single-tree search is used (as opposed to dual-tree
82 search.
83 --verbose (-v) [bool]
85 Display informational messages and the full list of parameters
86 and timers at the end of execution.
87 --version (-V) [bool]
89 Display the version of mlpack.
90
92 --indices_file (-i) [string] Output matrix of indices. Default value
93 ''. --kernels_file (-p) [string] Output matrix of kernels. Default
94 value ''.
95 --output_model_file (-M) [unknown]
97 Output for FastMKS model. Default value ''.
98
100 For further information, including relevant papers, citations, and the‐
101 ory, consult the documentation found at http://www.mlpack.org or
102 included with your distribution of mlpack.
103mlpack-3.0.4 21 February 2019 mlpack_fastmks(1)