Creating a Chatbot
What is a chatbot? A chatbot is an assistant for answering questions, saving time from waiting human鈥檚 responses. But how? How computer can understand human language? Indeed, it can鈥檛 read words like we do. It reads number. Therefore, how to transform words into numbers is crucial for using languages to communicate with computers. Data Preprocessing Depending on what kind of chatbot to build, the collection of data can range from open source platform, scraping from websites, accessing to APIs from data vendor. It may includes labels for doing classification problems. If not, having some domain knowledge is important to do proper labelling. ...
Recommender Systems and Applications of Berlin Grocery Shopping
As amount of data explosively increasing nowadays, we have got so many information to digest. On the one hand it is healthy growth of society for the diversity of information, but on the other hand having too many options distracts our attention and perhaps it is not efficient to obtain the knowledge we need. In order to filter out the most relevant information, recommender systems are adopted. Recommender systems are filtering models for recommended information to people. Generally they are categorized into three types. ...
Building a MA Crossing Strategy and Backtesting with Backtrader and VectorBT
While on the road of learning algorithmic trading, backtesting provides a safe and efficient way to examine the performance of strategies. In this article, I would like to share my experience of testing a trend following strategy through different approaches. A smooth mean average(SMA) crossover strategy is implemented on the historical data of BTC-BUSD asset. The trading logic is when the fast line of price crossing the slow line of price from below, a buy signal is on. Vice versa, the sell signal is on when fast one crossing slow one from above. 10 day鈥檚 average price(sma10) is used as fast line, and 20 day鈥檚 average price(sma20) is for slow line. ...