Anomaly Score Plot Calculator

Score observations using robust statistical methods and flexible thresholds. Inspect ranked anomalies with dynamic charts. Compare detection behavior before trusting production model alerts confidently.

Calculator Inputs

Supplied scores assume larger values are more anomalous.
Used only by fixed threshold mode.
Used only by percentile mode.
Ties may flag more than this count.
Example: 0.05 expects five percent anomalies.
Use one to disable moving average smoothing.
Chart Options
Separate values with commas, spaces, semicolons, or new lines.
Separate labels using commas, semicolons, or new lines.

Formula Used

Robust modified Z-score
Score = 0.6745 × |x − median(x)| ÷ MAD

MAD equals the median absolute deviation from the sample median.

Standard Z-score
Score = |x − mean(x)| ÷ standard deviation(x)

Directional modes keep only high or low deviations.

IQR fence distance
Score = distance beyond Q1 or Q3 ÷ (Q3 − Q1)

Values inside the quartile range receive zero distance.

Threshold rules
Flag = score ≥ threshold

The comparator can require strict threshold exceedance instead.

How to Use This Calculator

  1. Choose raw values or supplied anomaly scores.
  2. Select a robust, standard, quartile, or bounded method.
  3. Choose two-sided, high-only, or low-only detection.
  4. Set a fixed, percentile, count, or contamination threshold.
  5. Paste values and optional matching labels.
  6. Adjust chart, smoothing, precision, and display options.
  7. Submit the form and review ranked flagged observations.

Understanding Anomaly Scores

Anomaly scores quantify how unusual each observation appears within data. Higher scores usually indicate deviation from expected behavior. Their meaning depends on the selected scoring method.

Scores support fraud checks, equipment monitoring, and quality control. They also help prioritize records for human review. A plot reveals patterns hidden inside isolated numeric values.

Why Plotting Scores Matters

Visual inspection shows clusters, spikes, drifts, and repeated alerts. These shapes often expose operational changes or scoring instability. Tables alone rarely communicate temporal behavior with equal clarity.

A threshold line separates routine cases from review candidates. Markers highlight flagged points without hiding surrounding score context. Smoothing can reveal sustained changes behind noisy local variation.

Choosing a Scoring Method

Robust modified scores resist distortion from extreme observations. Standard scores work well when distributions remain approximately symmetric. IQR distance focuses directly on values beyond quartile fences.

MAD ratios measure distance from the median using robust spread. Min-max scoring creates bounded values for simple visual comparison. Supplied scores preserve outputs from an existing detection model.

Threshold Design

Fixed thresholds apply stable domain rules across repeated datasets. Percentile thresholds adapt to each submitted score distribution. Top-count thresholds return a controlled number of review candidates.

Contamination estimates express the expected proportion of unusual observations. Smaller contamination values produce stricter, more selective alert boundaries. Threshold choices should reflect review capacity and error costs.

Interpreting the Plot

Single spikes may represent rare events, noise, or data errors. Consecutive spikes can signal drift, incidents, or changed processes. Gradual score increases may precede threshold crossings and failures.

Compare anomaly markers with original values and meaningful labels. Review the highest rankings before investigating borderline threshold cases. Context remains essential because unusual does not always mean harmful.

Using Diagnostics Responsibly

Check score averages, dispersion, alert rates, and maximum values. Unexpected alert growth may indicate drift or poor calibration. Very low alert rates can also hide missed anomalies.

Test several methods before deploying a monitoring rule. Compare flagged records against verified outcomes whenever labels exist. Recalibrate thresholds after data or model changes.

Common Machine Learning Uses

Security teams score logins, transactions, devices, and network activity. Manufacturers monitor vibration, temperature, pressure, and production measurements. Data teams inspect pipelines for corruption and schema shifts.

Practical Evaluation

Measure precision when false alerts create expensive investigation workloads. Measure recall when missed anomalies carry serious operational consequences. Review both metrics when tradeoffs remain uncertain or changing.

Final Guidance

Anomaly scores guide attention rather than proving abnormal intent. Reliable decisions combine statistics, context, review, and documented policies. Careful monitoring keeps detection useful as conditions evolve continuously.

Frequently Asked Questions

1. What does an anomaly score represent?

It represents relative unusualness within a dataset or model. Larger scores generally indicate stronger deviation. The exact scale depends on the selected method.

2. Which scoring method handles outliers best?

Robust modified Z-scores usually handle existing extremes better. They use medians and median absolute deviation. Standard scores can shift when severe outliers affect averages.

3. How should I choose a threshold?

Use verified outcomes when they exist. Otherwise, compare percentile and contamination settings carefully. Choose a level matching review capacity and operational risk.

4. What does contamination mean?

Contamination estimates the expected anomaly proportion. A value of 0.05 expects roughly five percent. It creates a threshold from the upper score distribution.

5. Can I enter model-generated scores?

Yes, choose the supplied score input mode. Enter scores in their existing observation order. Larger supplied scores are treated as more anomalous.

6. Why can top count flag extra records?

Several observations may share the boundary score. Inclusive comparison flags every tied observation. Select strict comparison to exclude exact boundary ties.

7. Does smoothing change anomaly classification?

No, smoothing only adds a visual moving average. Flags use the original calculated scores. This preserves exact threshold decisions and rankings.

8. Are high scores always harmful?

No, unusual observations can be valid or beneficial. Scores only prioritize records for investigation. Domain review determines whether an event needs action.

9. How many values can the calculator process?

The calculator accepts up to five thousand values. Large charts may require more browser memory. Use meaningful labels only when they improve interpretation.

Related Calculators

Time-Series Anomaly PlotControl ChartIsolation Forest Decision PlotReconstruction Error HistogramNormal-versus-Anomalous Scatter PlotThreshold Selection Curve

Important Note: All the Calculators listed in this site are for educational purpose only and we do not guarentee the accuracy of results. Please do consult with other sources as well.