Adding upstream version 1.32.0.upstream/1.32.0

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 55 insertions, 0 deletions

diff --git a/ml/kmeans/KMeans.cc b/ml/kmeans/KMeans.cc
new file mode 100644
index 000000000..e66c66c16
--- /dev/null
+++ b/ml/kmeans/KMeans.cc
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+#include "KMeans.h"
+#include <dlib/clustering.h>
+
+void KMeans::train(SamplesBuffer &SB, size_t MaxIterations) {
+    std::vector<DSample> Samples = SB.preprocess();
+
+    MinDist = std::numeric_limits<CalculatedNumber>::max();
+    MaxDist = std::numeric_limits<CalculatedNumber>::min();
+
+    {
+        std::lock_guard<std::mutex> Lock(Mutex);
+
+        ClusterCenters.clear();
+
+        dlib::pick_initial_centers(NumClusters, ClusterCenters, Samples);
+        dlib::find_clusters_using_kmeans(Samples, ClusterCenters, MaxIterations);
+
+        for (const auto &S : Samples) {
+            CalculatedNumber MeanDist = 0.0;
+
+            for (const auto &KMCenter : ClusterCenters)
+                MeanDist += dlib::length(KMCenter - S);
+
+            MeanDist /= NumClusters;
+
+            if (MeanDist < MinDist)
+                MinDist = MeanDist;
+
+            if (MeanDist > MaxDist)
+                MaxDist = MeanDist;
+        }
+    }
+}
+
+CalculatedNumber KMeans::anomalyScore(SamplesBuffer &SB) {
+    std::vector<DSample> DSamples = SB.preprocess();
+
+    std::unique_lock<std::mutex> Lock(Mutex, std::defer_lock);
+    if (!Lock.try_lock())
+        return std::numeric_limits<CalculatedNumber>::quiet_NaN();
+
+    CalculatedNumber MeanDist = 0.0;
+    for (const auto &CC: ClusterCenters)
+        MeanDist += dlib::length(CC - DSamples.back());
+
+    MeanDist /= NumClusters;
+
+    if (MaxDist == MinDist)
+        return 0.0;
+
+    CalculatedNumber AnomalyScore = 100.0 * std::abs((MeanDist - MinDist) / (MaxDist - MinDist));
+    return (AnomalyScore > 100.0) ? 100.0 : AnomalyScore;
+}