# Regular aeon Developer Meeting (early) - May 05, 2023
###### tags: `aeon-regular-dev-meeting`

#### May 05, 2023 @ 12:00AM (noon) UTC
#### Previous meeting: https://hackmd.io/Ph4gFbZzS9KBCMu3Bcv5Aw?view

Present: AJB, MM, AF, CH, LT

### Topics

- [Tony] Governance model
- Currently it is an adaption from sktime. A much simpler model has been proposed
- see https://github.com/aeon-toolkit/aeon/pull/195
- Maybe we should replace the CC role with individual role for core developers

- [Tony] Grant resources
- Money for pydata/hackathons/interns. Can probably move between.
- 3x12 week summer interns. £12k
- Hackathons/code sprints. £21k
- Pydata and conference presentations. £7.5k

Interns. Maybe wait until next summer, or split into shorter periods?
We should prioritise pydata, and identify projects/people/events

Project ideas:
1. time series machine learning with deep learning
2. distance based time series machine learning (when port in PF2)
3. time series classification
4. forecasting deep learning
5. lancaster forecasting
6. annotation

Events:
Ali: Paris. Dates?
Guilherme: Brazil.
Many: London
Amsterdam Call for proposals will close on Sunday, June 11, 2023

global
london
paris (January 2024)
madrid
brazil (where?)
Proposals:
1. Deep learning
2. Probabilistic

- [Tony] Benchmarking Interface
- See https://github.com/aeon-toolkit/aeon/pull/379 does this work for classification?

Possible topics for later meetings:
- [Tony] Transformer Interface
- See https://github.com/aeon-toolkit/aeon/pull/263
- [Tony] Unequal length series
- progress, next steps

### Need decision

N/A

### Need attention (reviews)

Comment on the current governance PR (https://github.com/aeon-toolkit/aeon/pull/195)
Kotsu PR (https://github.com/aeon-toolkit/aeon/pull/379)

### Previous Action items

N/A

### Action items

1. Read governance PR, comment, approve, request changes and have a core dev vote if necessary. Communicate with NumFOCUS after if possible for ideas.
2. Write a couple of projects for interns (with lancaster). Forecasting and annotation/outlier detection.

Ideas:
a) deep learning for forecasting
b) lancaster forecasting
c) annotation (changepoint, anomaly, outlier detection)

3. Plan a core dev meeting for the summer if possible
4. Put together a proposal for Pydata Amsterdam for Chris (and one other) for the distances module
5. Ask Markus for Pydata global proposal
6. Form a list of meetings and times for Pydata events and other possible talks
7. Read the benchmarking PR, decide if it can be used for machine learning use cases

### Next meeting date

May 19, 2023 @ 17:00 UTC