The peak of deaths is expected to occur between April 11 and April 14, but until then, the number of deaths per day will remain relatively static, between 700 and 900.
At this point in the pandemic, the UK has confirmed 6,159 deaths from coronavirus, of which 786 were recorded on Tuesday.
The country is said to be at least a week behind Spain and Italy, which have registered more than 10,000 deaths but are now decreasing in daily deaths.
The Telegraph model, compiled by our data science team, examines the latest data from nine different countries – China, Japan, South Korea, Spain, Germany, France, Italy, the United States and the United Kingdom – assess the spread of the coronavirus.
Then, after taking into account other factors such as population demographics, hospital resources, test levels and government social distancing measures, we are able to estimate where the UK is located on its own epidemiological curve and the time and height of its own peak.
While deaths are expected to peak in about a week, we are currently in the peak of the number of cases, depending on the model. Indeed, if a person were to die from a coronavirus, it usually takes a week for their symptoms to be present.
The model estimates that the UK is currently experiencing its peak in new cases. A total of 3,634 confirmed cases were recorded on Tuesday, bringing the total to 55,242.
Our model estimates that it will start to fall, two weeks after Boris Johnson, the Prime Minister, announced the British foreclosure.
Another model, compiled by Imperial College London, suggested that around 20,000 people in the UK were dying from the virus. This takes into account the lockdown imposed by the United Kingdom, with the same modeling predicting that deaths could reach a quarter of a million if social distancing were not applied.
A note on modeling
The Telegraph analysis consists of an overall machine learning model that ingests Covid-19 case and death data over time and generates a forecast of the number of cases and deaths by country for the next 30 days.
Our machine learning model is constantly trained on all Covid-19 data available in the nine countries mentioned above. As a result, the model refines its 30-day forecast window daily.
Numbers and dates should only be viewed as a window into which the peak could fall, rather than a specific number. It is difficult to predict the spread of a pandemic, with consequences such as a nursing home escape having major implications.
Therefore, our analysis is only intended to serve as a guide to show the approximate time in which the peak occurs, and how expensive it could be.
The members of the Telegraph data science team who produced the model are Davoud Ardali, Juliana Maia, Yves Jacquot and Jack Bennett.