asavinov

7 ms·

1.
by asavinov 2y ago
You could check out a tool for trade signal generation based on machine learning and feature engineering: https://github.com/asavinov/intelligent-trading-bot It trains ML models based on historic data and custom featur
2.
by asavinov 3y ago
The sequence FROM Table AS t WHERE t.Condition SELECT t.col1, t.col2, ... might be more natural than the traditional SELECT t.col1, t.col2, ... FROM Table AS t WHERE t.Condition If we compare it with how loop are described
3.
by asavinov 3y ago
From my experience with crypto currencies and the intelligent trading bot [0] I would say that transformers will not provide significant benefits when applied to the traditional statistical (numeric) forecasting problems. Such models assume
4.
by asavinov 3y ago
I agree that the conventional (numeric) forecasting can hardly benefit from the newest approaches like transformers and LLMs. I made such a conclusion while working on the intelligent trading bot [0] by experimenting with many ML algorithms
5.
by asavinov 3y ago
Does it work synchronously or asynchronously? For example, if you integrate WebSockets then you could act as soon as you get new data. If it works synchronously then you regularly request data (say, every second) and then act. If you proces
6.
by asavinov 3y ago
When developing an automatic trading system the following aspects are important: - Data feeds and data ingestion. It can be a fairly independent component which collects data from different sources (might be even discussion forums) making i
7.
by asavinov 3y ago
> What gives you advantage is trading algo, which is always hard to find. At the end it is necessary to make a decision whether to buy or sell (and how much), which will compete with other decisions made based on some logic. Developing
8.
by asavinov 4y ago
For any such tool, two questions are of primary importance: - How connections between multiple tables are represented and managed - How derived data is described (queries, workflows etc.) Typically such tools are aimed at simplifying data
9.
by asavinov 4y ago
> I have always kept in mind is that feature engineering is almost always the key difference between success and failure I also developed an ML-powered service heavily relying on feature engineering https://github.com/a
10.
by asavinov 4y ago
> Joins are what makes relational modeling interesting! It is the central part of RM which is difficult to model using other methods and which requires high expertise in non-trivial use cases. One alternative to how multiple tables can
11.
by asavinov 4y ago
One idea is to use columns instead of cells. Each column has a definition in terms of other columns which might also be defined in terms of other columns. If you change value(s) in some source column then these changes will propagate throug
12.
by asavinov 5y ago
Looks interesting although I miss a short introduction or how-to guide. I found that one can "Create and train machine learning models to predict market values." In this context, a related project is Intelligent Trading Bot: http
13.
by asavinov 5y ago
Having Python expressions within a declarative language is a really good idea because we can combine low level logic of computations of values with high level logic of set processing. A similar approach is implemented in the Prosto data pro
14.
by asavinov 5y ago
Similar products: - https://rows.com/ - https://www.rowshare.com/ - https://www.airtable.com/ - fieldbook - Power BI
15.
by asavinov 5y ago
Most of the self-service or no-code BI, ETL, data wrangling tools are am aware of (like airtable, fieldbook, rowshare, Power BI etc.) were thought of as a replacement for Excel: working with tables should be as easily as working with spread
16.
by asavinov 5y ago
Hamilton is more similar to the Prosto data processing toolkit which also relies on column operations defined via Python functions: https://github.com/asavinov/prosto However, Prosto allows for data processing via colu
17.
Show HN: Intelligent Trading Bot using machine learning and feature engineering(github.com)
13 points by asavinov 5y ago|1 comments
18.
by asavinov 5y ago
> I think SQL is irritatingly non-composable, many operations require gymnastics to express One approach to radically simplify operations with data is to use mathematical functions (in addition to mathematical sets) which is implemen
19.
by asavinov 5y ago
One alternative to SQL (type of thinking) is Column-SQL [1] which is based on a new data model. This model is relies on two equal constructs: sets (tables) and functions (columns). It is opposed to the relational algebra which is based
20.
by asavinov 5y ago
Here is another project based on the same idea of processing data using functions : https://github.com/asavinov/lambdo - Feature engineering and machine learning: together at last! Yet, here the focus is on feature en
21.
by asavinov 5y ago
The main motivation is that the conventional approaches to data processing are based on manipulating mathematical sets for all kinds of use cases: we produce a new set if we want to calculate a new attribute, we produce a new set if want
22.
Show HN: Functions matter – an alternative to SQL and map-reduce data processing(github.com)
6 points by asavinov 5y ago|2 comments
23.
by asavinov 5y ago
> I always felt like there was some super deep & fundamental link between these mathematical concepts and relational modeling ideas. The relational model is relies on set theory (more specifically relational algebra). An alternative
24.
by asavinov 5y ago
> This is closer to the way that humans perceive the world — mapping between whatever aspect of external reality you are interested in and the data model is an order of magnitude easier than with relational databases. One approach to m
25.
by asavinov 6y ago
Assume you need to execute the following query: SELECT SUM(A) FROM MyTable For large tables it will take some time to compute the result. Now assume we append a new record and want to get the new result. The traditional approach is
26.
by asavinov 6y ago
Yes, this is the basic logic: for any incremental aggregation we need to detect groups which can be influenced by this new record or updated record. If we do row-based rolling aggregation then then indeed we need to update records (i-n, i+n
27.
by asavinov 6y ago
Having a possibility to update (query) output with new input data rather than process the whole input again even if the changes are very small is indeed a very useful feature. Assume that you have one huge input table and you computed the
28.
by asavinov 6y ago
I find this project quite interesting because sklearn has a good general design including data transformations and it does make sense to provide compatible functionality for Go. Feature engineering in general is a hot topic and especially i
29.
by asavinov 6y ago
You write that it is "based on sklearn". Does it rely on sklearn functions and requires Python?
30.
by asavinov 6y ago
The real power of such kind of tools is determined by its ability to derive (infer) new data from multiple tables, particularly, by linking tables and aggregating data (using these links). Airtable had an interesting approach to this prob

More ›