MLHub Currated Packages

The MLHub repository hosts the following currated pre-built machine learning models. Try them out and let us know if you have any issues. They are easily and quickly installed and demonstrated. Feedback is welcome through github. Visit MLHub.ai for details.

Anyone can create a MLHub package simply by including a MLHUB.yaml file in their github repository. The models listed here are currated in by MLHub administrators. If you find any issues do be sure to report them.

Catalogue


Name Version Description
animate 2.1.5 Tell a data narative through animations
audit 4.1.0 Classic financial audit predictive classification model.
azanomaly 3.1.4 Azure Anomaly Detection.
azcv 2.7.2 Azure Computer Vision.
azface 2.2.1 Azure Face API demo.
azlang 0.0.4 Azure language cognitive service on the cloud.
azspeech 4.4.1 Azure Speech cognitive services on the cloud.
aztext 2.5.2 Azure Text Analytics cognitive services on the cloud.
aztranslate 2.5.3 Azure Text Translation cognitive services on the cloud.
barchart 2.0.2 Demonstrate the concept of barcharts.
beeswarm 2.0.1 Demonstrate the concept of bee swarm charts.
bing 0.1.5 Bing Maps
cars 1.0.0 Identify car make and model from a photo.
colorize 1.5.9 Demonstrate the concept of photo colorization.
cvbp 2.2.0 Computer vision best practices.
deepspeech 0.0.3 Deepspeech
easyocr 0.0.8 Extract text from images.
facedetect 0.2.5 Simple face detection.
facematch 0.4.2 Simple face recognition.
google 0.0.1 Google Maps
iris 2.1.3 Classic iris plant species classifier.
kmeans 0.3.0 An animation demonstration for the kmeans clustering
movies 2.0.4 Movie recommendation using the SAR algorthm.
objects 1.6.27 Recognise objects in an image using resnet152.
ocsvm 0.0.5 Introducing one-class support vector machine.
opencv 1.0.3 OpenCV Computer Vision.
patientpaths 0.0.8 Report patient paths for specific scenarios.
ports 2.0.2 Demostrate the concept of visualising data.
pyiris 0.0.8 Classification models in Python using the iris dataset.
pyspeech 0.1.3 Convert audio speech to text across multiple services.
rain 5.1.4 Predict if it will rain tomorrow (decision tree and rand...
rbm 1.0.6 Recommendations using restricted Boltzmann machine.
sar 1.1.6 Smart adaptive recommendations.
scatter 2.0.1 Demonstrate the concept of scatter plots.
sgnc 0.1.1 Node classification for graphs using StellarGraph.
tapwater 0.0.3 Factor analysis for understanding customers
webcam 1.1.0 Capture video, process, feed dummy device for Zoom.
zynlp 0.0.11 Tweets sentiment analysis.

Showcase

Visualisation

Visual presetnation of data is crucial for understanding data and sharing the data story with others. Animations can be quite effective in telling the story over time. The basic visualisations include the barchart and the scatter plot. A more informative scatter plot is the beeswarm plot. Visualisations are effectively used in all kinds of reports as in this study of Australian sea ports.

Prediction

Machine learning algorithms are typically deployed for the task of prediction and classification. The rain package, from Rattle, includes a model to predict if it will rain tomorrow using decision tree and random forest algorithms. The audit package, also from Rattle, demonstrates the task of identifying clients who should be audited for tax compliance, for example. The traditional example used to demonstration classification is the iris package using decision trees to predict the iris plant species. A variation on this using Python is available as pyiris. Recommendation systems also perform prediction to be able to suggest movies to watch. A one class support vector machine (ocsvm) can be used to identify outliers.

Computer Vision

carsobjectsopencvcolorizefacedetect

Azure Cloud AI

azcvaztextaztranslateazfaceazspeechazanomalyazlang

Index

academic publications

adult

animation

anomaly detection

athletics

australian government

azure

bar chart

best practices

bing maps

biology

blurry

brand detection

cars

cart

classification

climate

cloud

cluster

colorize

command detect

command similar

computer vision

consumer involvement

convolution network

cora

decision tree

deep learning

deep neural networks

deepspeech

dnn

entity

face detection

face recognition

factor analysis

geocode

ggplot2

google

google maps

graph

graph convolution network

graphics

health

image classification

image to text

imagenet

introductory

iris

k nearest neighbours

keras

kmeans

language

line chart

linear discriminant analysis

location

map

motor vehicles

movie recommendation

naive bayes

natural language processing

natural language understanding

neural network

object detection

ocr

opencv

outliers

patient pathways

phrases

plot

policy

prediction

psych

python

python3

pytorch

r

random data

random forest

recommendation

resnet

rpart

sar

scatter plot

sea ports

sentiment analysis

spatial

speech

speech synthesis

speech to text

sports

stellar graph

support vector machine

tensorflow

text analytics

text to speech

text translation

thumbnail

transcription

translation

video

visualisation

weather

webcam

zoom