|
1 | 1 | # Changes in the BigML.io Java bindings
|
2 | 2 |
|
| 3 | +## 1.4 version |
| 4 | + |
| 5 | +### New Features |
| 6 | + |
| 7 | +* Added support for Local Anomalies |
| 8 | + |
| 9 | + A new LocalAnomaly class to score anomalies in a dataset locally or embedded |
| 10 | + into your application without needing to send requests to BigML.io. |
| 11 | + |
| 12 | + This class cannot only save you a few credits, but also enormously reduce the |
| 13 | + latency for each prediction and let you use your models offline. |
| 14 | + |
| 15 | + Example usage (assuming that you have previously set up the BIGML_USERNAME |
| 16 | + and BIGML_API_KEY environment variables and that you own the model/id below): |
| 17 | + |
| 18 | + // API client |
| 19 | + BigMLClient api = BigMLClient.getInstance(); |
| 20 | + |
| 21 | + // Retrieve a remote anomaly by id |
| 22 | + JSONObject jsonAnomaly = api.getAnomaly("anomaly/551aa203af447f5484000ec0"); |
| 23 | + |
| 24 | + // A lightweight wrapper around a Anomaly resurce |
| 25 | + LocalAnomaly localAnomaly = new LocalAnomaly(jsonAnomaly); |
| 26 | + |
| 27 | + // Input data |
| 28 | + JSONObject inputData = (JSONObject) JSONValue.parse("{\"src_bytes\": 350}"); |
| 29 | + |
| 30 | + // Calculate score |
| 31 | + localAnomaly.score(inputData); |
| 32 | + |
| 33 | + |
| 34 | +* Added support for Samples resources. |
| 35 | + |
| 36 | + Full API documentation on the API can be found from BigML at: |
| 37 | + https://bigml.com/developers/samples |
| 38 | + |
| 39 | +* Added support for Project resources. |
| 40 | + |
| 41 | + Full API documentation on the API can be found from BigML at: |
| 42 | + https://bigml.com/developers/project |
| 43 | + |
| 44 | +* Added support for predictions using Median. |
| 45 | + |
| 46 | +### Improvements |
| 47 | + |
| 48 | +* We improve the predictions responses adding the more information: Distribution, |
| 49 | + DistributionUnit, Count and Median values associated to the prediction. |
| 50 | + |
| 51 | +* Added methods in LocalModel class to get information of the model. Added methods: Summarize, |
| 52 | + Data Distribution and Prediction Distribution. |
| 53 | + |
| 54 | +* Added BaseModel and ModelFields base classes to reuse logic in LocalCluster and LocalPredictiveModel |
| 55 | + |
3 | 56 | ## 1.3 version
|
4 | 57 |
|
5 | 58 | ### New Features
|
|
0 commit comments