1 /* ----------------------------------------------------------------------
2 * Copyright (C) 2019-2020 ARM Limited. All rights reserved.
3 *
4 * $Date: 09. December 2019
5 * $Revision: V1.0.0
6 *
7 * Project: CMSIS DSP Library
8 * Title: arm_svm_example_f32.c
9 *
10 * Description: Example code demonstrating how to use SVM functions.
11 *
12 * Target Processor: Cortex-M/Cortex-A
13 *
14 * Redistribution and use in source and binary forms, with or without
15 * modification, are permitted provided that the following conditions
16 * are met:
17 * - Redistributions of source code must retain the above copyright
18 * notice, this list of conditions and the following disclaimer.
19 * - Redistributions in binary form must reproduce the above copyright
20 * notice, this list of conditions and the following disclaimer in
21 * the documentation and/or other materials provided with the
22 * distribution.
23 * - Neither the name of ARM LIMITED nor the names of its contributors
24 * may be used to endorse or promote products derived from this
25 * software without specific prior written permission.
26 *
27 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
28 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
29 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
30 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
31 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
32 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
33 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
34 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
35 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
37 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
38 * POSSIBILITY OF SUCH DAMAGE.
39 * -------------------------------------------------------------------- */
40
41 /**
42 * @addtogroup groupExamples
43 * @{
44 *
45 * @defgroup SVMExample SVM Example
46 *
47 * \par Description:
48 * \par
49 * Demonstrates the use of SVM functions. It is complementing the tutorial
50 * about classical ML with CMSIS-DSP and python scikit-learn:
51 * https://developer.arm.com/solutions/machine-learning-on-arm/developer-material/how-to-guides/implement-classical-ml-with-arm-cmsis-dsp-libraries
52 *
53 * \example arm_svm_example_f32.c
54 *
55 * @} */
56
57 #include <math.h>
58 #include <stdio.h>
59 #include "arm_math.h"
60
61 /*
62 The polynomial SVM instance containing all parameters.
63 Those parameters can be generated with the python library scikit-learn.
64 */
65 arm_svm_polynomial_instance_f32 params;
66
67 /*
68 Parameters generated by a training of the SVM classifier
69 using scikit-learn and some random input data.
70 */
71 #define NB_SUPPORT_VECTORS 11
72
73 /*
74 Dimension of the vector space. A vector is your feature.
75 It could, for instance, be the pixels of a picture or the FFT of a signal.
76 */
77 #define VECTOR_DIMENSION 2
78
79 const float32_t dualCoefficients[NB_SUPPORT_VECTORS]={-0.01628988f, -0.0971605f,
80 -0.02707579f, 0.0249406f, 0.00223095f, 0.04117345f,
81 0.0262687f, 0.00800358f, 0.00581823f, 0.02346904f, 0.00862162f}; /* Dual coefficients */
82
83 const float32_t supportVectors[NB_SUPPORT_VECTORS*VECTOR_DIMENSION]={ 1.2510991f, 0.47782799f,
84 -0.32711859f, -1.49880648f, -0.08905047f, 1.31907242f,
85 1.14059333f, 2.63443767f, -2.62561524f, 1.02120701f,
86 -1.2361353f, -2.53145187f,
87 2.28308122f, -1.58185875f, 2.73955981f, 0.35759327f,
88 0.56662986f, 2.79702016f,
89 -2.51380816f, 1.29295364f, -0.56658669f, -2.81944734f}; /* Support vectors */
90
91 /*
92 Class A is identified with value 0.
93 Class B is identified with value 1.
94
95 This array is used by the SVM functions to do a conversion and ease the comparison
96 with the Python code where different values could be used.
97 */
98 const int32_t classes[2]={0,1};
99
100
main(void)101 int32_t main(void)
102 {
103 /* Array of input data */
104 float32_t in[VECTOR_DIMENSION];
105
106 /* Result of the classifier */
107 int32_t result;
108
109
110 /*
111 Initialization of the SVM instance parameters.
112 Additional parameters (intercept, degree, coef0 and gamma) are also coming from Python.
113 */
114 arm_svm_polynomial_init_f32(¶ms,
115 NB_SUPPORT_VECTORS,
116 VECTOR_DIMENSION,
117 -1.661719f, /* Intercept */
118 dualCoefficients,
119 supportVectors,
120 classes,
121 3, /* degree */
122 1.100000f, /* Coef0 */
123 0.500000f /* Gamma */
124 );
125
126
127 /*
128 Input data.
129 It is corresponding to a point inside the first class.
130 */
131 in[0] = 0.4f;
132 in[1] = 0.1f;
133
134 arm_svm_polynomial_predict_f32(¶ms, in, &result);
135
136 /* Result should be 0 : First class */
137 #if defined(SEMIHOSTING)
138 printf("Result = %d\n", result);
139 #endif
140
141 /*
142 This input vector is corresponding to a point inside the second class.
143 */
144 in[0] = 3.0f;
145 in[1] = 0.0f;
146
147 arm_svm_polynomial_predict_f32(¶ms, in, &result);
148
149 /* Result should be 1 : Second class */
150 #if defined(SEMIHOSTING)
151 printf("Result = %d\n", result);
152 #endif
153
154 #if !defined(SEMIHOSTING)
155 while (1); /* main function does not return */
156 #endif
157 }
158
159
160
161
